Q. Why do we need audit our elections? Why don’t we just fix them?

A. Many appropriate technical solutions have been offered by reformers but, crucially, those solutions require legislative approvals and vigorous enforcement by the very officials who most benefit from the status quo. They are not likely to reform their systems voluntarily; they may give in if pushed to the wall by voter revolts or criminal investigations. This goes for Democratic and Republican political machines alike.

And the federal government is of little help. In testimony July 24 before the House Oversight Committee, Thomas Hicks, one of the two Commissioners on the Election Assistance Commission, said that the Commission could not certify new (and presumably better) election systems because it lacked a quorum – the President having serendipitously failed to fill the two remaining seats. Solutions requiring action by the federal government are therefore D.O.A. until the FEAC board has a quorum.

Q. Isn’t gerrymandering a bigger problem for American democracy?

A. Counting errors are not the whole story; the principal Republican bulwark against unfavorable rising demographics is gerrymandering (Democrats engage in gerrymandering as well, though to a lesser extent because they control fewer state legislatures). Unfortunately, judicial challenges to gerrymandering are likely to become more difficult with Anthony Kennedy’s retirement, so we are probably stuck with partisan gerrymandering for some time. That means that we have to work harder on fixing problems that don’t depend on the Supreme Court for action: eliminating election fraud and attacking voter suppression.

Q. Don’t states already audit their elections? Why do we need another layer of auditing?

A. Most states already have some kind of “audit” process, but it is usually run at the county level by officials responsible to the local ruling party. The audits, therefore, in corrupt systems can be caused to validate the official results. (Here is an example from Cook County, Illinois.) Even in “clean” systems there is no guarantee that such audits will reveal official errors, not to mention crimes. And even when crime is suspected courts have been reluctant to act without direct evidence of error or fraud — which, if criminal, will be well-concealed by the malefactors. It has been all but impossible, therefore, to take effective legal action within the election certification time frame.

So election results must be audited using a different technology or technique than that used to generate the original count, conducted independently of party machines, and done quickly so as to allow forensic investigation and corrective action before the “winning” candidate takes office.

Q. What’s the point of counting correctly? Voting doesn’t matter anyway. That’s why many Americans don’t vote.

A. American voter turnout is low for a variety of reasons. Many people are cynical about the differences between the parties (“It doesn’t matter who I vote for, the result is always the same”), others say one vote makes an infinitesimal difference (which is true, but when groups vote together their votes make a big difference), and some believe their votes are not counted and elections are stolen, so why bother?

We’re working on that last problem. If there is a good chance that fraud will be exposed, that everyone’s votes will be added up, and that illegitimate winners will be prevented from taking office, many in the last two groups will turn out to vote. This will have an impact on who takes office, which in turn will make the parties more sensitive to the voters in that first group.

Q. Why does counting the popular vote matter so much? The Electoral College chooses the U.S. President anyway, even if the other candidate wins the popular vote.

A. In 2000, 2004, and 2016, the popular vote was extremely close in the states that gave the winners their Electoral College victories. In all three elections, machine errors, incompetence, fraud, and voter suppression may have put them over the top in those states. In such close elections any errors potentially can change the Electoral College winner.

Consider the math:

In 2000 Bush’s winning margin in Florida, the state that put him over the top, was 500-odd votes, but the Secretary of State there had systematically removed more than 50,000 likely Democratic voters from the rolls so they couldn’t vote. The Secretary of State was a co-chair of Bush’s election campaign.

In 2004 Kerry was projected by the polls and by exit polls to win by around a million votes in Ohio, but after a “computer crash” in the Secretary of State’s office the results started reversing and Bush won by about 100,000 votes. Later investigations turned up evidence pointing to computerized fraud. Again, the Secretary of State was a co-chair of Bush’s election campaign.

In 2016 the President won Michigan by two-tenths of one percent, yet a Michigan election administrator in December 2017 stated that Michigan’s vote tabulators have a demonstrated error rate of between 0.26% and 1.78%, and “have miscounted ballots over 3%”[i] Add in tampering with ballot boxes and the existence of 75,000 ballots lacking a vote for President[ii] and the margin of error, potentially fraudulently induced, grows even larger.

In Wisconsin the margin of victory was 27,252 votes, but a study using highly conservative assumptions estimated that in just the two counties studied Wisconsin’s new voter I.D. law prevented and deterred at least 16,801 people (disproportionally minorities), “and could be as high as 23,252”.[iii]

In Pennsylvania, where the victory margin was about eight-tenths of a percent, 80 percent of voters vote on ancient and demonstrably vulnerable touchscreen voting machines that do not keep a paper record and that the Department of Justice reported problems with.[iv] The winning margin was an average 6 votes per touchscreen precinct, an easily achievable number. Jill Stein sued to examine the machines, a request that was denied by a Pennsylvania court on multiple grounds.[v]

In other words, strictly speaking, the President’s margin of victory may not have been legitimate; at the very least it is mathematically impossible to know the legitimate winner of the Electoral College and therefore the Presidential election.

2016 was not unique. Fraud, error and voter suppression have been common in American elections for generations, and the spread of computerized voting equipment have only made errors more difficult to catch and easier and less risky to cause. In three of the last five elections the Republican presidential candidate has lost the popular vote but won the Electoral College vote by a razor-thin margin. It is entirely possible that this will happen again in 2020, irrespective of whom the Republican candidate is. Why? Because the election systems in the swing states that decide the Electoral College are error-prone, highly vulnerable to tampering and manipulation, and are run by Republicans committed to the reelection of the Republican candidate.

Most of America’s 8,000+ voting systems are vulnerable to manipulation; the lack of transparency and verifiability can hide error and reward fraud. Credible statistical analyses have demonstrated bizarre phenomena for which fraud is the most probable explanation.

Q. It sounds like you think only Republicans commit election fraud. The Democrats are no angels either. 

A. Absolutely not. In the not-so-distant past the Democratic machines had the reputation for the most election fraud — because they controlled more governments. It’s very possible that John Kennedy would have lost the election to Richard Nixon in 1960 but for Democratic fraud in Chicago and Texas. There was a lot of circumstantial and statistical evidence that smelled like fraud in the Democratic primaries in 2016 between Hillary Clinton and Bernie Sanders. And the primary race that year between Cong. Debbie Wassermann Schultz and her challenger Tim Canova revealed some extremely suspicious issues.

The apparent pattern is of party machines protecting themselves against challengers, whether those challengers are from the other party (during general elections) or their own party (during primary elections). Whichever party is in power in any locale pretty much determines who will be the perpetrator of fraud if there is any. Politics is a blood sport, and the Golden Rule of politics is “do unto others as they’ve done unto you, and do it twice as hard”.

At the moment Republicans control more state and county governments, so they have more opportunities to commit fraud. It therefore stands to reason that they will be doing more of it, even if they are no more underhanded than Democrats on a per-capita basis. If and when Democrats start picking up more states and counties they are likely to go back to their old ways. To prevent this, and to clean up the system so neither party can steal elections with impunity, we need to expose and root out fraud now, without fear or favor, and in the ensuing political storm impose institutional reforms that make fraud much harder to commit.

Q. What are we as a country doing to protect our elections?

A. Very little in many states: The institutional barriers to prosecution of election fraud are pervasive.

The officials who run elections are usually partisan office holders who often exhibit party loyalties that appear to affect their official functions; Secretaries of State in the crucial swing states of Florida and Ohio in 2000 and 2004 served as co-chairs of their party’s presidential campaigns, a clear conflict of interest that is not barred by law.

When required audits of election results are conducted by the same officials who run suspicious elections those officials possess an incentive to conceal their actions. After the 2016 Democratic primary in Illinois, credible observers testified to the Cook County Board of Elections that officials and staff overtly conformed discrepant audit to official results when the latter did not agree with the audit results.[vi]

Law enforcement investigations are rare; prosecutions are even more rare.

“Winners” take office long before evidence of fraud is acted on. And when prosecutions do occur the illegitimate victors are rarely removed from office.

Q. Why don’t courts intervene to control election fraud?

A. Unfortunately judicial action is also not very effective at present. The right to vote is not mentioned in or guaranteed by the federal Constitution – it is left to the States. Therefore, except when racial discrimination is demonstrated, voting and, derivatively, accurate counting to guarantee the effect of the vote, is not privileged under federal law. This means that judicial oversight of possibly wayward election officials is also left up to the States. In practical terms, therefore, transparency and accountability are partisan issues and may be non-existent, as was demonstrated in Michigan in 2016.[vii]

Furthermore, American courts typically regard statistical and other scientifically generated evidence as speculative. A scientist may be convinced by statistical significance at a 95% level, but courts see this as a 5 percent probability of being wrong – or more, if the opposing counsel attacks the underlying research methodology, which he usually will. Thus exit polls and other statistical tests and evidence of machine vulnerability do not meet most courts’ burden of proof as legally admissible evidence of misfeasance, at least for the case in chief. Independent direct evidence of suspicious discrepancies and fraud is therefore required. For instance, the Federal District Court for the Eastern District of Michigan, in the case of Green Party candidate Jill Stein’s request for a recount in Michigan, held that

“[T]o date, Plaintiffs have not presented evidence of tampering or mistake. Instead, they present speculative claims going to the vulnerability of the voting machinery — but not actual injury. Because mere potentiality does not amount to a claim that the vote was not fairly conducted, Plaintiffs’ new claims are insufficient to maintain the existing TRO. No likelihood of success on that claim has been shown.

The issues that Plaintiffs raise are serious indeed. The vulnerability of our system of voting poses the threat of a potentially devastating attack on the integrity of our election system. But invoking a court’s aid to remedy that problem in the manner Plaintiffs have chosen — seeking a recount as an audit of the election to test whether the vulnerability led to actual compromise of the voting system — has never been endorsed by any court, and would require, at a minimum, evidence of significant fraud or mistake — and not speculative fear of them. Such evidence has not been presented here.”[viii]

Under this sort of reasoning, courts generally deny discovery motions against election authorities, calling them “fishing expeditions”. This means that as long as corrupt authorities can keep evidence of fraud concealed, the courts will not permit inquiry.

Q. What should be the standard for election integrity?

A. The United States needs to take election fraud just as seriously as it does securities fraud, because the public interest in accurate vote counting is at least as great, if not greater than, financial fraud.

Here’s an example of the way it should be:

US law requires publicly owned companies to be independently audited every quarter, even when they are not suspected of any wrongdoing. Why? Because the 1933 Securities and Exchange Act postulated that there is a strong public interest in protecting shareholders and financial markets, and the best way to protect them is to require frequent independent oversight, on the theory that oversight increases transparency and transparency deters wrongdoing. And if wrongdoing still occurs, the Securities and Exchange Commission is poised to prosecute and send offenders away for long stretches.

Q. What about voter fraud? Isn’t it important? What are you doing about it?

A. “Voter” or “voting” fraud, i.e., individual voters who voted twice or impersonated someone else, is a crime with very little practical consequence, because the personal calculus discourages it happening more than occasionally: A voter who votes more than once contributes infinitesimally to the outcome but faces felony prosecution for his troubles. A Pew study in 2012 found no evidence of serious voter fraud,[ix] and in 2016 the author said that they had seen no evidence to support claims of massive voter fraud.[x] There are historical stories of voters being bused around voting multiple times, and people voting dead people, but because those are organized efforts, usually organized by a political machine of some kind, they more properly fall under the category of election fraud. Our system will detect these.

Q. What about voter suppression?

A. A similar conundrum applies to the problem of voter suppression. In cases involving discrimination on the bases of race, sex and ethnic origin, injury to one person is enough to guarantee standing to sue because those are protected categories under the US Constitution. In contrast, because voting is not a US Constitution-protected right, courts have rejected lawsuits alleging partisan voter suppression: They reason that the injury to the small number of plaintiffs does not rise to the level of a Constitutional injury and in any event their small numbers would not have affected the outcome of the election in question. This has made it very difficult for election activists to attack voter I.D. laws that successfully avoid being perceived by courts as motivated by discrimination against a protected category of persons.

Here is an article describing the suppressive effects in Wisconsin of severe voter I.D. laws.

And here is a brief article on how voter suppression is carried out in practical terms and how you as a voter can prevent being prevented from voting.

Q. What are other election integrity groups doing? Are they succeeding?

A. The strategy of most election integrity groups is to lobby legislatures to adopt transparent, verifiable voting and counting systems, return to paper ballots, stricter official audits, greater transparency, etc. These are all good technical solutions, but their chances of adoption are greatest where they are needed least. Where they are needed most the solutions require the consent of the very politicians who owe their seats to the working of unclean systems. They are not likely to voluntarily change the rules that won them their seats – they will have to be compelled.

The question is, how to compel them?



[i] Jan Ben Dor, Michigan Election Reform Alliance press conference, December 19, 2017.

[ii] Ibid.

[iii] Kenneth R. Mayer and Scott McDonnell, “Voter ID Study Shows Turnout Effects in 2016 Wisconsin Presidential Election”, press release summarizing findings at

[iv] The calculation of 6 votes per was done as follows: Pennsylvania has 9,156 precincts; 80% of 9,156=7,325. The difference between the two candidates was 44,292. 44292/7325=6 (rounded to whole numbers).

[v] JILL STEIN, et al., v. PEDRO A. CORTÉS, et al., accessed at

[vi] This video of the meeting of the Cook County Board of Election Commissioners contains credible descriptions of audit personnel conforming audit counts to apparently invented official results in the primary contest between Hillary Clinton and Bernie Sanders.

[vii] Jan Ben Dor in the press conference cited above cites numerous blatant failures of legal oversight viz overtly suspicious election official activities in the 2016 Presidential election.


[ix] The original report, Election Fraud, August 30, 2012, is at

[x] The Hill, no date shown,



Please support our work. Help restore integrity to elections and accountability to government. Click here to donate.