The Maricopa Audit: Separating Fact from Fiction

Author John Droz Jr @

These days citizens interested in election integrity are confronted with conflicting narratives. On the one hand they hear those who make frenzied claims of flagrant fraud. On the other hand others (like the media) insist that the election process is pristine. Separating the wheat from the chaff is a significant challenge.

Towards that end, let’s cut through the noise and get one fact crystal clear: The current Maricopa election audit is NOT a full forensic audit!

Acknowledging that reality does not diminish what is being done, as the Maricopa audit is a much better effort than what any other US county has done for the 2020 elections — or likely before. So kudos to the AZ legislators for making it happen!

To clearly understand exactly what is being done (and not done) in the Maricopa audit, we need to carefully examine the Statement of Work (SOW) — which is the official contract from the State of Arizona hiring a company to do the Maricopa election audit.

The SOW (like most employment contracts) is a reasonable compromise based on normal considerations like cost, time-frame, data available, etc.

Another interesting reality to be aware of is that the SOW was primarily written by the employee, not the employer. The main reason for that is that Maricopa County is going down unchartered waters, and the contractor is more knowledgeable about the ins-and-outs of election audits than the state of Arizona is.

To understand what is missing, we need to know what a Full Forensic Audit (FFA) actually is. This is explained in the Post-Elections Audits Assuring Election Integrity report. Basically, a FFA is a three-part comprehensive investigation, into each of the main potentially problematic voting areas: 1) Voters, 2) Machines, and 3) Process.

The SOW shows that the Maricopa audit does investigate each of these three key areas — but with limitations. These limitations mean that the Maricopa audit is a partial forensic audit, not a full one. Let’s look at an example to better understand this.

The first part of a FFA is a comprehensive analysis of voters. The best current example of how a full forensic audit of voters is done, is this report for the entire state of Nevada.

Attorney Jesse Binnall’s team of data analysts carefully compared public records to names (and addresses) of those who actually voted in Nevada in 2020. They didn’t have to call anyone (or knock on any doors), as this was an unobtrusive comparison of public information.

Their revealing findings for the 2020 elections were then publicly testified to (in front of a US Senate Committee, on national TV), under oath:

  • 1,500± dead people are recorded as voting.
  • 4,000± non-citizens voted.
  • 8,000± people voted from non-existent addresses.
  • 15,000± votes were cast from commercial or vacant lot addresses.
  • 19,000± people voted even though they did not live in Nevada.
  • 42,000± people voted more than once.


“All in all, our experts identified over 130,000 unique instances of voter fraud in Nevada. But the actual number is almost certainly higher. Our data scientists made these calculations not by estimations or statistical sampling, but by analyzing and comparing the list of actual voters with other lists, most of which are publicly available. Our evidence has never been refuted, only ignored.” [As a point of comparison, Trump lost Nevada in 2020 by 33,000± votes.]

Is this type of list comparison (which is an essential part of a full forensic audit) being done in the Maricopa audit? No.

Considering that Binnall’s team did the entire state of Nevada, doing the same analysis for just Maricopa county Arizona would likely be a similar effort. (Note: although Maricopa County has more people than Nevada, it is much more concentrated.)


What is being done regarding the Voter component in the Maricopa audit?


The answer is in the SOW. In Section 4.1 (page 3) it makes clear that the Voter analysis is based on statistical samplingnot a 100% list match of every person who voted (as was done for Nevada).  For example it says:

— “Complete audit of a minimum of three precincts”

[It seems that there are 750± Maricopa precincts. So this is a “complete audit” of three out of 750± precincts, which is a very small sample.]

— “Analysis of existing research and data validating the legitimacy of voter rolls.”

[Not clear what “analysis of existing research and data” etc. means, as the SOW does not elaborate on the specifics.]

— “Comparing results against known lists of invalid voters (e.g. deceased voters, non-citizens, etc.)”

[Note that this does not appear to include some of the disparities identified in the Binnall report — e.g.  non-existent addresses, non-residents, etc.]

In other words, the Maricopa audit Voter component falls short of being a full forensic analysis of possible voter irregularities.

A similar situation exists with the other two components: Machine and Process.

So how much of a Full Forensic Audit is being done in Maricopa? Since some of the details are not spelled out (as noted above), our guess is that this Partial Forensic Audit is about 75%± of what a Full Forensic Audit would consist of.

Again, none of this is intended to disparage or minimize the good being done with the Maricopa Audit. Their level of audit is beyond what anyone else is doing!

However, it’s extremely important to understand that if there are no substantial issues identified by the Maricopa audit, it can NOT be said that a: “full forensic audit was not able to uncover consequential malfeasance or irregularities.”

A more accurate statement would be: “this partial forensic audit was not able to uncover consequential malfeasance or irregularities.”

Additionally, if XYZ problems are identified, it can NOT be said that a: “full forensic audit only uncovered XYZ as malfeasance or irregularities.”

A more accurate statement would be: “this partial forensic audit was able to uncover XYZ as some examples of malfeasance or irregularities.”

We are in turbulent waters regarding US election integrity, and we definitely do not need more misinformation and misunderstandings here.

To avoid that we should be clear about terminology, and exactly what is being done at significant election-related events like the Maricopa audit.

Several other states are closely following what is happening with the Maricopa audit. They should improve on the good work being done there, not just try to replicate it.