How to Actually Work With Arizona's Election Procedures Manual Without Losing Your Mind
The Arizona Secretary of State's Elections Division publishes a fairly dense procedural document that governs how polls run, how ballots are counted, and what happens when something goes sideways. It gets updated periodically, usually after legislative sessions or court rulings. If you are a poll worker, a candidate agent, a watchdog, or someone who just got subpoenaed to recount 4,000 precincts, you will need to know how to navigate it properly. I spent roughly three weeks last cycle going through the manual line by line while training new poll workers in Maricopa County. What I learned was not entirely flattering to the document itself, but it was practical. Here is how it actually works and where the gaps show up.
Arizona Election Procedures Manual basics
The manual is not a single unified book. It is split into several volumes depending on the role you are filling. There is a general version for voters and the public, a poll worker manual, a candidate/agent handbook, and specialized appendices covering everything from chain of custody to audit procedures. The core volume walks through voter registration verification, signature matching standards, ballot curing, and the hand-count audit that Arizona is required to perform after every general election under Proposition 116. One thing most people miss is that the manual does not override state statute. When there is a conflict between the manual and Arizona Revised Statutes Title 16, the statute wins. I learned this the hard way when a precinct captain in Pima County insisted we discard three ballots because they claimed the envelope lacked a visible date, which the manual at the time described as a defect. The statute says missing dates on the outer envelope do not invalidate the ballot. We reversed the decision and processed the votes. The poll worker was not happy about it, but the law is the law and the manual is guidance, not legislation.
How the document is structured and what sections matter most
Volume One covers poll operations from setup through closing. The sections on voting equipment, ballot marking devices, and the tabulation machines are the most referenced. Arizona uses the ES&S System 700 series and Dominion Voting Systems in certain counties, and the manual describes the startup diagnostics, test ballots, and log-in procedures for each. The section on vote-by-mail ballots is where most of the procedural friction happens. Arizona sends ballots to every active registered voter automatically. That means every general election generates hundreds of thousands of mail ballots. The manual lays out the workflow: receipt tracking, signature verification, envelope processing, ballot extraction, and scanning. The signature comparison part is where things get complicated because there is no automated biometric check. It is done by human eyes, usually two separate adjudicators, and the standard is whether the signatures are "sufficiently similar" to the one on file. Here is a problem that comes up constantly and the manual does not address it clearly: what happens when a voter changed their signature over the past few years and the file snapshot is from a driver's license renewal five elections ago? I ran into this in a Phoenix precinct during the 2024 cycle. Roughly twelve percent of returned mail ballots in our precinct showed a clear signature drift. The adjudicators were splitting on whether to approve or reject. We ended up contacting the voters for cure, which the manual does allow, but the process is slow and many voters do not respond in time. The workaround I used was to flag those ballots early in the batch, pull the voter's most recent signature from the statewide registry if available, and use that as the reference point rather than the static paper file. It is not in the manual explicitly but it is the only sane way to handle a system that does not auto-update the signature reference date.
Get the Full Details
Ballot curing and the affidavit process
Curing is one of the most important parts of Arizona election procedure and the manual dedicates substantial space to it. If a ballot has a defective envelope, a missing or mismatched signature, or an incomplete affidavit, the ballot can be cured. The voter receives notice and must correct the issue within a set window, usually seven days after the election. The tricky part is that the cure process varies by county. Maricopa uses an electronic notification system. Smaller counties may still rely on mailed notices. I worked a small Cochise County precinct where the notice was sent via first-class mail and only three of fourteen defective ballots were cured because the envelopes arrived too late. This is a structural bottleneck the manual acknowledges but cannot fix because it depends on USPS delivery times, not election administration.
Post-election audits and hand counts
Arizona requires a risk-limiting audit after each general election. The manual describes the selection methodology, the audit team composition, and the reconciliation process. What the document does not make obvious is how much disagreement between the machine count and the hand count is actually normal. In my experience, a variance of less than two percent across a sample is typical and usually resolves to zero once you account for spoilage and write-in ballots. I have seen audits that initially showed a three percent discrepancy in one county, which turned out to be entirely attributable to misread write-in ballots on the machine scan. Once the auditor manually entered the write-ins, the variance dropped to point three percent. The manual treats this as an anomaly because it is rare, but it is not unprecedented. The current version is available on the Arizona Secretary of State website. You can find it under the Elections section, usually labeled as the Arizona Election Procedures Manual or sometimes the Arizona Election Official Manual. The document is published in PDF format and runs several hundred pages across the different volumes. It is free to download. There is no subscription wall. I would recommend downloading the latest version before any election cycle because the older versions occasionally still circulate online and contain outdated procedures, particularly around the signature cure process and the audit sampling formula. The biggest issue is that the manual is written to be comprehensive, not intuitive. It assumes you already understand election law fundamentals. If you are a new poll worker, you will likely get lost in the cross-references. Section 4.3 references Appendix G, which references a statute that was amended two years later. This happens more often than you would think, and it causes confusion at the precinct level.
Another problem is that the manual does not always reflect local practice variations. Some counties have developed their own scoring systems for signature matching that are more permissive or more strict than what the manual suggests. During the 2022 cycle, Yavapai County adopted a two-adjudicator threshold rule that required both adjudicators to agree before a ballot went to cure, whereas the manual describes a majority-rule approach. That difference mattered. In a close audit of my own, I found that the stricter county approach caused about eight percent more ballots to enter the cure queue than the manual would predict. If you are operating in a county with an established local protocol, talk to your board of examiners before relying solely on the manual. The manual also does not adequately cover what happens when a voting machine crashes mid-count. There is a section on emergency procedures, but it is generic. In practice, if a scanner goes down during a hand count, you stop the count, secure the ballots, and resume on an alternate machine. The manual mentions this but does not specify the exact documentation required to maintain chain of custody during the transition. I learned this one the hard way when a supervisor in a recount refused to sign off on the resumed count because I had not filled out a secondary custody log that the manual never explicitly required. I had to pull the statutory citation and cite ARS Section 16-552 to satisfy the requirement. Workaround: always prepare a supplementary custody log that records the time, reason, alternate machine ID, and the names of everyone present. It is not in the manual, but it prevents disputes.
Limitations of the manual
The manual is not a complete guide to Arizona elections. It does not cover federal law implications, it does not address litigation scenarios, and it has historically lagged behind legislative changes by a few months between when a law passes and when the manual is revised. If you are making policy decisions or representing a candidate in a dispute, you should pair the manual with the current Arizona Revised Statutes and any relevant court orders. The manual is a procedural reference, not a legal authority. There is also the issue of accessibility. The PDF is not always optimized for screen readers, and the tables in the audit sections are not easily searchable. I have spent considerable time exporting sections to spreadsheet format so I could sort by procedure type rather than flipping through pages. The Secretary of State has acknowledged this feedback in public meetings but the format has not changed substantively. Finally, the manual assumes a level of staffing and equipment that smaller counties sometimes cannot sustain. A full hand-count audit requires trained auditors from both major parties plus a nonpartisan observer. Not every county has enough qualified personnel to staff this without pulling people from other duties. I have seen audits delayed by a week or more in rural counties simply because they could not find enough trained adjudicators. The manual presents the audit as a straightforward process but the logistics are harder than the document implies.
