the nevada independent logo
Clark County Election Department personnel process mail in ballots on Friday, Oct. 30, 2020. (Jeff Scheid/The Nevada Independent)

Regardless of your political party, can I ask you to dig into some data with me, with an open and scientific mindset, setting aside who won and who lost? There is a bigger issue at hand than just this election: whether we will ever trust the legitimacy of a vote that our side (whichever it is) loses. To draw from Ronald Reagan’s “shining city upon a hill” parable, I suggest we consider America that bright city on a hill — the eyes of all people are upon us as we demonstrate what it means to be a democracy.

I was enlisted by “The Trump Digital Army, Election Integrity Division” on November 8th, 2020 to look at the source data in each battleground state, and compare 2016 to 2020 results to find any outliers/anomalies in the 2020 election results that could signal fraud. Here is the data they asked me to assemble:

  • Voter registration (2016 vs. 2020)
  • Turnout by county (2016 vs. 2020)
  • Turnout exceeding registered voters
  • Margin of victory by county, Republicans vs. Democrats  (2016 and 2020)
  • Age anomalies in voter registration (e.g. potentially dead voters or under-age voters)
  • Out of state voters (and party registration)

The source data for every Nevada analysis is available to the public via the secretary of state and the Clark County registrar of voters. I’ve posted all my source file links in the report I provided to the Trump Digital Army, Election Integrity Division.

Findings

  1. Voter registration favored Republicans. The largest change in registration was in pro-Trump counties, such as Pershing, Mineral and Lincoln. For example, in Lincoln, Democrats lost 7 percentage points while registered Republicans increased by 5 percentage points.
    • Finding: There was no evidence of rigging the voter registration process for Biden in Nevada.
  1. Turnout by county was higher overall, with the largest increases in pro-Trump rural counties.
    • Finding: There was no evidence of rigging turnout for Biden in Nevada.
  1. Turnout exceeding registered voters would indicate ballot-stuffing; however, there is no county where the number of ballots cast exceeded the number of registered voters. The three highest voter turnout figures are from Humboldt, Douglas and Eureka, all counties that pulled heavily for Trump.
    • Finding: There was no evidence of stuffing ballots for Biden in Nevada.
  1. The margin of victory in 2020 was only 0.3 percent different from 2016.

The table below follows the Trump Digital Army, Voter Integrity format. Red indicates Republican-leaning, and blue is Democratic-leaning. The numbers are the point margins in 2016 and 2020, as well as the change from 2016 to 2020. The last (white) column is the percent for each county of total votes cast in the state in 2020. (Clark and Washoe represented 86.9 percent of all votes in Nevada.)

Democrats actually lost ground in votes in Clark County, the main source of Democratic support in the state, slipping from 10.7 percent margin in 2016 to 10.0 percent in 2020. This makes Clark County an unlikely source of systemic widespread fraud by Democrats.

Several rural counties increased the margin for Republicans (Lincoln, Humboldt, Esmeralda) at levels that represent potential outliers — however, I did not investigate outliers that favored President Trump.

A few rural counties, while still favoring Trump, did so a little less this election when compared to 2016 (Carson, Douglas, Churchill, Nye).

In 2016, Washoe went Democratic by 2,671 votes for a 1.3 percentage point margin. In 2020, Washoe went Dempcratic by 11,368 votes, or 4.6 percentage point margin — which makes Washoe an outlier for Biden.

    • Finding: Additional scrutiny of Washoe County voting patterns was warranted.

Washoe County analysis

Washoe delivered a more than 11,000-vote margin for Biden — a meaningful increase over the 2016 vote margin. I examined the raw voter files from Nov. 5 and Nov. 11 to look for evidence of fraud. I traced ballots for individuals, investigated Dominion software pre- and post-voting audits, and looked into other registration numbers and voting patterns.

  1. Age anomalies. On behalf of the Trump Digital Army, I examined two possibilities: Dead people voting and underage people voting.

The Trump Digital Army considered it a warning flag if there was a meaningful increase in older voter registrations, hinting at a coordinated campaign to register dead people and vote for them. I performed an age analysis of the voter files, looking at the share of registered Republicans and registered Democrats within the 65+ population as well as conducting a date of birth analysis.

There was virtually no change from 2016 to 2020 in registered Republicans (41.7 percent to 41.4 percent), and there was a slight down-tick in registered Democrats (40.2 percent to 38.7 percent). Given that the vote-by-mail file is drawn from the active voter roll, it is possible that some deceased persons may have been mailed a ballot, but it could not have amounted to many, given the way the voter file is checked against the Electronic Death Registry System (EDRS). There were no anomalies detected that would suggest widespread voter fraud.

Re: the voter record details in Washoe, there were a handful of anomalies in the data of birth dates, but this amounted to fewer than 30 out of more than 300,000 ballots cast (or, 1/100th of a percent). A few of the 30 appeared to be unreasonably old (e.g. DOB = 1900). Three were under 18 years of age. I called several of these voters; errors appeared to be clerical.

Finding: There was no evidence of coordinated voter fraud enlisting dead or underage voters.

  1. Mail-in ballots. There were accusations that fake ballots were printed and fed into machines to rack up votes for Biden. In investigating this possibility, I found no evidence for this theory, and several good data points indicating that such a theory of ballot-stuffing was not feasible.

Voted ballots must match voter records. The voting record contains the names and addresses of every voter in Washoe, along with a code for “Mail” indicating when a ballot was sent out, and a tag of “VBM” (along with a date) indicating when the ballot was received back.

My wife voted by mail, and I traced her ballot through the voter file. I could see a return date of 10/13/20. In contrast, I voted early (code “EV”) and noted my South Valley location (code “SOVL”), where I voted, and the voting timestamp of 10/23/20 at 12:35 hours. Here’s a snapshot of the vote trace:

Each person who voted, whether in-person or by mail, has a voter record consisting of their name, voter ID number, and a vote method code. Manufacturing fraudulent ballots would require creating fraudulent voter registration records before the election, stealing an active voter’s ballot and casting it or creating duplicate mail-in ballots. However, as already noted, there were no anomalies in registration levels pre-election. There also was no instance where ballots cast exceeding registered votes. Washoe didn’t even have the highest voter participation rate in the state (that honor goes to Humboldt County, with 89.2 percent). There were also no cases of duplicate mail-in ballots.

As each mail-in ballot comes with a uniquely coded envelope to match to the specific registered voter, the paper ballot is specialized and not easily forged. Further, the vote-counting process was live-streamed, and also observed by members of both parties, with an adjudication board that includes both Republican and Democratic Party representatives to handle any ballots that were un-readable by the machine.

Finding: There was no evidence of fraud via mail-in ballots.

  1. Out of state voters. Could someone rig the Nevada election by having lots of out-of-state Californians register to vote here?

Digging into this possibility, we find that about 3 percent of voters in the voter file have an out-of-state mailing address.

There are people who own more than one home and who consider Washoe to be their primary residence but who received a mail-in ballot out of state. There also was the potential for people who had moved and no longer resided in Washoe to be mailed a ballot (if they had changed their address, but did not indicate that they had moved out of state, for instance). The top three states for out-of-state mailing addresses in the voter file are: California, Arizona and Texas — a mix of left, center and right-leaning states.

More Republicans have their mail sent to them out of state (2,832) vs. Democrats (2,517), but it is relatively close — registered Republicans are 12 percent more likely to have an out-of-state mailing address.

Finding: There was no evidence of systemic fraud by one party trying to use out- of-state residents to skew the election. To the extent that there are illegal votes by people who live out of state, it is slightly more likely these would have favored President Trump, given that there are more registered Republicans vs. Democrats with out-of-state mailing addresses.

  1. Duplicates & double-voting (i.e., voting by mail and then in-person):

The Washoe voter file from Nov. 5 has 303,989 records; about 3,800 are duplicates. Some duplicates exist because of a change of address, as the following vote trace data shows (for a friend of ours who is building a new home).

They had changed their address at the DMV (which is linked to their voter file) to the one for the new home, but as the home was not completed in the run-up to the election, when ballots were mailed out to everyone, theirs could not be delivered to that address. The family showed up to early vote, and was asked to fix their address by the poll worker. You can see the challenge code “MAIL” in their record, meaning the mail-in ballot was undeliverable. You can also see an Early Vote Code for the ballot that was counted. You can also see the Voter ID duplicated as there are now two addresses associated with the voter. This type of duplicate record is considered a change of address.

continued...

Another type of duplicate record occurs because a person voted by mail, then voted again in person. Analysis shows that registered Democrats and Republicans have about the same rate of these duplicate vote entries.

In an effort to trace one of the duplicate votes, I interviewed a person whose mail-in ballot was collected from her home, but who was concerned that the person wouldn’t turn it in. She wanted to be sure her vote was counted, and appeared at a polling place in person, explaining the situation to the poll worker. In the file, I could see that the mail ballot did arrive — and that her in-person vote canceled out the vote by mail (as it should be).

Finding: No evidence of fraud. (Members of the Tea Party NV party were twice as likely to attempt double voting, however, it was still less than 50 people.)

  1. Ballot challenges

There were about 1,500 vote challenges logged in the Washoe Nov. 5 voter file. Challenges that are not cured result in a vote being thrown out. Of the challenges, 566 were for the votes of registered Republicans vs. 622 for registered Democrats. Registered Republicans were least likely to have their vote challenged. Democrats were 10.5 percent more likely to have their ballot challenged — though the challenge rate is still quite low (1 in 171, or 0.58 percent). “No Party” fared the worst; members were 42 percent more likely to have a vote challenged compared to registered Republicans, with 531 non-mail challenges, or 0.7 percent (1 in 133).

Finding: No evidence of systemic fraud via duplicate voting or vote challenges. If anything, the ballot challenges favored the Republican Party.

  1. Dominion software issues

I investigated the pre-election machine testing and the verification process that confirms that what is entered matches the vote tallies. I also checked a machine when I voted in-person and could examine the printed tape. I also checked the methodology of the post- election audit of the electronic vote tally vs. the tape count. All of them aligned, and all of those processes were open to the public and had both parties observe the audits.

When talking with a Republican poll observer, he commented, “There wasn’t much to observe because all they did was plug in the data key and a count shows up.” His lack of appreciation for the fact that the voting machines are air-gapped (not connected to the internet by any means) and how the pre- and post- election audits work to ensure integrity was concerning to me. I’d recommend continuity in poll observers from pre- through post- election audits, and a checklist to confirm that voting machines are not connected to the internet, that the printed tapes match the machine entries, and that the total entries match the total paper counts.

Finding: There was no evidence of machine hacking or fraud.

  1. Citizenship

According to the UCSIS.gov website, 843,593 people were naturalized in the U.S. last year. Over the past decade, more than 8 million people have gained their citizenship. Given Nevada’s population relative to that of the U.S., the math works out to about 8,500 new citizens in Nevada each year.

There was a claim that roughly 3,100 documented non-citizens in Nevada voted. This claim was based on matching voter records to DMV records and looking for a particular type of driver’s license that is typically issued only to documented non-citizens. However, because such a license can be good for up to four years, it is possible (in fact probable) that most of the 3,100 people with that special license became citizens prior to voting. After all, over the most recent four-year period, more than 30,000 people in Nevada would have become citizens.

There is no analysis possible to determine whether the claimed 3,100 were more likely to be registered Democrats or Republicans such that it would suggest systematic vote-rigging. While there certainly could be a few that voted illegally, it seems a big risk to jeopardize one’s ability to gain U.S. citizenship.

Finding: There is no requirement that upon naturalization, a person needs to have their license reissued. The data is inconclusive, and would require matching the 3,100 names to the voters’ files to confirm that they cast a ballot (as opposed to merely being in the voter rolls), with which party they were registered, and to match names against current naturalized citizens list to see how many (if any) non-citizens voted.

Conclusions

As in any process involving people, there will be mistakes. There will even be a few people who either through ignorance or deceit will vote more than once. (Nevada participates in a 30-state consortium to catch people who vote in more than one state — there are a few hundred nationwide each election.)

But I found nothing in the data that shows the election was rigged.

If anything, there are a few small factors that tilt slightly in favor of Republicans. Yet, overall trust in elections has plummeted among Republicans. According to Morning Consult polls prior to the election, 66 percent of GOP voters said they had at least some trust in the U.S. election system. In the latest poll (Dec. 15th), that number dropped to 33 percent. Democratic trust, meanwhile, jumped from 63 percent to 83 percent.

(If it wasn’t fraud, why did President Trump lose Nevada? See Part II of my deep dive into the Nevada election data, coming Wednesday.)

While I can’t argue that the system isn’t transparent (I’ve performed all this analysis and interviews based on publicly available data, and public access), I do think the system lacks visibility. The difference between visibility and transparency is the degree to which the ordinary voter is provided visual proactive updates in a manner that is easy to digest.

Frankly, as an ordinary consumer, I get better tracking and visibility in monitoring a delivery from Amazon than I do from something significantly more important — my vote. It is only because I applied my analytic training and put in many hours of work that I could rule out widespread fraud and increase my own trust in our election process. Perhaps it shouldn’t be so time-consuming or difficult for voters to see the data which addresses doubt about voter fraud.

Increasing visibility and faith in our democratic system of electing our representatives is in everyone’s best interest. It is my view that our state should make a concerted effort to better map the citizen’s voting journey, including clear updates for the individual’s vote, as well as a view of the overall patterns for the county — including one-tap access to videos of the voting and counting processes.

A bipartisan team should reach agreement on enhanced and digitized citizen observation of the voting process. This will increase visibility so there is greater trust in the system and less room for conspiracy theories to grow. It won’t convince those who will believe what they want to believe regardless of the facts, but it will help a meaningful percent of citizens who would like the reassurance of tracking their vote — and being able to see that careful checks for anomalies were conducted in order to rule out fraud and contextualize clerical mistakes.

Rex Briggs is an internationally recognized expert and published author in data analysis. He is the Founder and Executive Chairman of Marketing Evolution, Inc. He is a member of Research World’s advisory board, and was Wired’s first director of research. Briggs holds a bachelor of science degree from California Polytechnic State University, where he now serves on their Business Analytics Advisory Board. He resides in Nevada and is a registered nonpartisan. He contributed to both Republican and Democratic candidates in the 2020 election cycle. On Twitter: @rexbriggs

Comment Policy (updated 10/4/19): Please keep your comments civil. We reserve the right to delete comments or ban users who engage in personal attacks, use an excess of profanity, make verifiably false statements or are otherwise nasty.

What Happened Here: A six-part series on COVID-19 in Nevada

correct us
ideas & story tips