Definitions, Questions and Source Information
Definitions
Definitions under Post-Election Day Processes
Canvassing is the comprehensive process of collecting and matching ballot data, and of reviewing election procedures, that is a critical part of ensuring the accuracy of results and improving election administration from year to year. It does everything from matching in-person voter lists (“poll books”) with ballots cast, and the number of mail ballots sent with those received, which allows for tabulating the final, official election results; to noting and investigating discrepancies and hiccups in the process. As the Election Assistance Commission describes it, the canvass is “a culmination of all the data generated during an election cycle.” The members of canvassing authorities vary state by state.
Certification is the process of finalizing election results, in which election officials attest to their accuracy and finality, and notify candidates of their victory. The certifying officials vary state by state.
Audits, or in this context “post-election audits,” are reviews of election results and equipment to ensure accuracy and functionality. The National Conference of State Legislatures categorizes most audits as either traditional or risk-limiting, and both of these audits is a tabulation audit — meaning it is a review of actual votes, not just equipment or processes.
A traditional audit “look[s] at a fixed percentage or set number of voting districts or voting machines and compare[s] the paper record to the results produced by the voting system.” This is the most common type of audit.
A risk-limiting audit adjusts the percentage of reviewed ballots based on the margin of victory in a given race; generally put, the larger the margin, the fewer the ballots necessary to audit, and the more that election officials’ resources can go toward greater needs, including audits of tighter races. (Both the number of the margin of victory and the number of total votes cast in a given race are relevant to determining the “risk” that a certain number of audited votes found to have been miscounted could affect the outcome of an election.)
Definitions under State Mail Ballot Deadlines
Postmarks are the official stamps of the U.S. Postal Service that date when a piece of mail was officially accepted into circulation. For the purposes of elections, they are relevant in 14 states that count mail-in ballots which are postmarked by Election Day but arrive at election facilities after Election Day, by a period of time that varies by state. (Even more states have such “grace periods” for only military and overseas voters.) The legality of these “postmark rules” was upheld by the Supreme Court in the case Watson v. Republican National Committee. Declare’s summary of the decision is here, and Declare contributor Ben Ginsberg’s analysis of the case is here.
Definitions under Voter ID Laws
“Mostly mail” elections are elections in which all registered voters are mailed a ballot, with opportunities that vary by state to then vote by mail or still vote in person. In mostly mail states, most voters choose the option to vote by mail. There are eight such states (alphabetically): California, Colorado, Hawaii, Nevada, Oregon, Utah, Vermont, and Washington. Because “voter ID” typically refers to something shown in person, it is less relevant in mostly mail states, and the process of voter verification weighs more heavily on the time of registration, voter list maintenance, and, in some states, the inclusion of certain personally identifying information with mail ballots.
Signature matching is the process of matching a voter’s signature on file with the government (on a voter registration form, for example) with a signature on a ballot or a poll book. For the purposes of demonstrating in-person voter ID, it is one option for in-person voters who fail to provide the requested form of identification at the polling place.
Questions
This map contains a lot of data, about 50 states, and from separate sources. How confident are you that it’s all 100 percent accurate?
This would be an entirely fair question that needs to be answered humbly. First, issuing corrections is simply part of a responsible media enterprise — and should a reader have doubts that a particular piece of data in this map is incorrect or misrepresented, we would want to know about it, so that we could investigate and set the record straight, if needed. info@declarenews.org is the address to email.
Second, we are confident and secure in the process that we used to construct this map. First, the National Association of Secretaries of State, the National Conference of State Legislatures, and the Election Assistance Commission are all authoritative data sources. Ultimately, the 50 state election codes and communications of state election officials are the “primary sources” relevant to this project. But we consider as qualified translators of the first category the official association of the country’s chief election officials (NASS); the official association of all 50 states’ legislators (NCSL); and a technical federal agency (EAC). We also believe that analyzing and excerpting statute in a handful of cases of vagueness, and using it to explain how we interpret such cases, is sufficient.
We hope to refine and become more precise, as warranted, over time. And the thing about a project like this is that it can actually help prompt the political process itself to refine and become more precise. We take as an example Delaware, where a reporter for Spotlight Delaware was ping-ponged among official entities involved in the state’s post-election process when trying to explain how it works. Laws themselves are only as authoritative as they’re clear and have a shared interpretation.
Why the specificity of “2026 Midterms” in the map title?
In some states, not every election is certified by the same state authority. In Arkansas, for example, the governor certifies federal elections and the state assembly certifies state ones. In Georgia, the secretary of state certifies congressional elections and the governor certifies presidential electors. Then, there’s the possible need to differentiate between the certification of presidential electors to Congress under the Electoral Count Reform Act, and certifying results under state laws. There’s no reason to disentangle all this right now, when the election processes of most interest and debate concern congressional races. And so the post-election processes you see on this map are specifically for congressional elections — even if many of them apply to presidential ones, too.
Why limit the noncitizen voting data to state audits?
They’re the most illustrative examples we have of the “scope” of the problem. There are other factoids about noncitizen voting, such as the number of states that have contracted with the federal government to use its SAVE tool to help verify voter citizenship, or that use other data-sharing arrangements, like in Georgia. In late June, however, a federal judge ruled that the Trump administration couldn’t allow the states to use SAVE, which Congress created as a tool for determining benefit eligibility for certain classes of immigrants, for the new purpose of determining voter eligibility. Ultimately, states’ noncitizen voting audits together provide the largest set of publicly available data, and function similarly to the Heritage Foundation’s Election Fraud Map.
As you can see, many, but fewer than half of states have shared the findings of a noncitizen voting audit this decade. For states with no audit listed on the map, we chose “n/a” as a descriptor — because we cannot say for certain if a state hasn’t conducted such an audit, or if it has and hasn’t released the information.
What’s up with all these different categories of voter ID?
The National Conference of State Legislatures sorts voter ID laws into four categories: “strict” requirements for photo ID or non-photo ID, and “non-strict” requirements for photo ID or non-photo ID. This kind of categorization is necessary, because the states with voter ID laws — 36 of 50 — have all sorts of different standards.
We use slightly different terminology, however, because even some “strict” voter ID states have exceptions, no matter how narrow — typically in cases of avowed indigence (cost-prohibitive poverty) or religious objection to being photographed. We find that it’s easier to draw the line at “no exceptions” versus “with exceptions,” and listing those exceptions by state in the map.
Source Information
Information about post-Election Day processes, state mail ballot deadlines, and voter ID laws was compiled from the National Association of Secretaries of State, the National Conference of State Legislatures (1, 2, 3), and the Election Assistance Commission. The information was compared with a similar data set published in 2022 by Michigan State University professor Quinn Yeargain at Bolts, for the purpose of adding an additional check and with the understanding that some of the information could be out of date. Finally, in instances where the NASS, NCSL, and EAC information was vague or conflicting, or it conflicted with Professor Yeargain’s work, Declare editor Chris Deaton read the relevant states’ election statutes and made a determination of how to list and, where deemed appropriate, footnote the data. Chris has an established record of interpreting the fine print of appropriations, healthcare, and immigration legislation and policy, and has been recognized for his fairness by the likes of Social Security and Medicare trustees.
Information about non-citizen voting audits was compiled from manual searches of primary sources by Declare contributor Stephen Richer, who is a senior fellow at the Harvard Kennedy School, Cato Institute legal fellow, and former Maricopa County recorder.