Throughout the course of the 2016 Democratic presidential primaries, one of the most well-documented and widespread forms of election fraud was registration tampering. We define registration tampering as the alteration of a voter’s party affiliation or registration status without their knowledge or consent. The outcome may be:
- the voter’s party affiliation is changed to a different party or to no party
- the voter’s status is changed to permanent vote-by-mail from in-person voter
- the voter’s status is not translated to the correct polling site, meaning they are kept off of voter rolls despite being listed as an active voter
- the voter’s status is changed to unregistered
After widespread reports of registration tampering in Arizona’s 2016 presidential primary election, Election Justice USA began collecting testimony from voters across the United States who had identified or experienced similar problems. At the same time, Election Justice USA collected explicit evidence corroborating these testimonies in the form of scans and screenshots of registration records, telephone call logs, and e-mails from elections officials.
All in all, Election Justice USA received verified reports of registration tampering from nearly 20 states.
Some of the earliest stories date back to the Fall of 2015. Election Justice USA‘s databases include four such reports from Colorado, which held a closed caucus on March 1, 2016. Among the most detailed is that of Alan Jennings, who sent the following detailed report and has agreed to be named and pictured here in conjunction with his story:
I was a registered voter in Colorado and had voted in prior elections. In about April/May of 2015, I went online to the voter registration website to change my affiliation from “unaffiliated” to “Democrat” following news that Bernie Sanders was going to run under the Democratic ticket. I had changed my voter affiliation to unaffiliated after the mid-terms in 2014.
I made the change, and thought nothing about it until late in October 2015, when I checked it again from hearing in social media that some voters in the state had mysteriously had their affiliations changed against their wishes. Sure enough, mine had been changed back to “unaffiliated” again. I changed it back to “Democrat,” and in November, around Thanksgiving checked it again, and it was still showing “Democrat.”
In Colorado, a closed caucus state, the deadline for making changes to your registration is January 4th. I checked it one more time in late December, and found it to still be set to “Democrat.”
However, in mid January, I read a disturbing post on Facebook in a Colorado group supporting Bernie Sanders that voters’ registration affiliations were still changing mysteriously, so I went back online to check mine, and found once again that it had been changed back to “unaffiliated.” I corrected it again, called the county and inquired about why it was changed and they acted as if they had no idea. So I escalated it to the State level, contacting the Attorney General and threatening notification of news media.
I received a phone call from the AG, telling me he had looked into my registration and seen my prior visits to the website, and that I had “not changed anything.” I told him my changes that I had had to make twice to change it back to Democrat, and he said he saw that I had been on the site, but that no changes had been made. I complained about the last change being past the deadline to vote and he said he would get that corrected, which he did, and in a subsequent email, it showed the date of registration as 9/4/2014.
That said, in posts to the social media site, one of the members of that Colorado Bernie Sanders group stated that he looked me up on the voter rolls and that I was not listed. I asked him whether that list came from the county or the state, and told him about my corrected registration, and he said neither, that the voter rolls were provided by NGP VAN, the company that installed the data breach on the Democratic National Committee’s servers three times in the fall.
When I went to the Caucus, I was not on the rolls, and had to fill out an affidavit to vote in them. Since it was a Caucus my vote WAS counted, but I am sure there were many turned away that hadn’t gone through all the steps I had and didn’t have the needed documentation.
By January, according to a follow-up discussion with Jennings, Democrats with access to NGP-VAN, the Democratic Party’s “get-out-the-vote” computer system, were noticing that registration switching was a widespread problem. “[W]hen I spoke with a precinct captain online [in January] who had access to the voter rolls, he said that I wasn’t on them, and would have to fill out an affidavit to participate. … In talking to him further, he indicated that the problem was widespread in closed primary states and closed caucus states, and that many voters’ affiliations were being switched.”
Unlike Mr. Jennings, who caught the issue in time and battled his way to being able to vote, most such voters were denied the right to vote with the proper ballot for their candidates of choice or, frequently, were forced to employ provisional ballots which are not always counted in many parts of the country. Victims of registration tampering were entirely disenfranchised if their provisional ballots traced back to clerically or maliciously deleted or altered registration records.
The iconic case of voter registration tampering in the library of Election Justice USA’s election irregularities was reported by the lead plaintiff of our first lawsuit, Campanello, et al. vs. New York State Board of Elections, et al.1
Leonard Campanello anticipated voting in the New York Presidential Preference Election of 2016. But, Arizona’s disastrous March presidential primary created viral accounts of systemic voter disenfranchisement on all channels of social media and, with growing numbers of wary New Yorkers reporting their own stories of being erased from voter rolls, Leonard checked his registration status through the State DMV website.
Campanello found that his record still existed but his party affiliation had been switched from Democrat to Republican by way of a “change of party affiliation” form bearing his signature. According to the time stamp on the official record, this change was recorded about six months after the plaintiff’s last actual interaction with the DMV (the reissuance of his misplaced driver license) in 2014. As discussed below, hard proof exists that fraudulent changes to party affiliation registration status were often back-dated, sometimes by half a decade or more.
After the Suffolk County Board of Elections sent a copy of his registration form to Campanello, it became clear that the forger of the document had replicated a “pixel-by-pixel” identical copy of his driver’s license signature (electronically affixed by the DMV to every one of his licenses issued since 2008) onto the form that re-registered him as a Republican voter.
Forged signatures reported to Election Justice USA have not always been copied so identically. Alba Guerrero’s story, relayed in the executive summary of this report is an example. Guerrero’s clearly forged signature helped convince a judge in New York City to allow her to vote. Guerrero’s false registration as Republican was back-dated to 2004, and she told Election Justice USA by phone that the Board of Election worker at the election day court hearing only had the two documents with unmatched signatures to represent her interactions with the registration system. All other information, to the extent it existed, may have been scrubbed by the forger.
Meanwhile, since first registering to vote in 2007, Campanello had been a Democrat and voted that way ever since. The discovery of his involuntary Republican affiliation was made well after the October 9, 2015 deadline to change parties and this fact prohibited Campanello from voting for his chosen candidate in New York’s April 19th, 2016 presidential primary. That candidate was to be Senator Bernie Sanders.
The documented forgery of voter signatures was not a common report of complainants reporting registration tampering, but that fact may only reflect the low likelihood that complainants would go to great lengths to discover the cause of their altered registration status, or that forgers who used this tactic early on discovered a more efficient way to tamper with registrations. Voter registration and election administration probably epitomize the meaning of “government bureaucracy” to the average voter who may be more likely to attribute computerized record problems to clerical error than to malicious design. Leonard Campanello’s case was not unique in New York (we have at least two other reports matching this description) or the country, with at least two reports from California involving faked signatures in switched party affiliations. It is not difficult to find other cases of forged registration signatures reported online.
Election Justice USA received complaints from 30 states and territories of the United States. From 18 of those states, reports offer a picture of voter registration systems in the world’s most technologically advanced country that are, at best, partisan and sloppy. At worst, these systems involve shoddy administrative and security protocols that cannot safeguard voter information against internal or external challenges nor reproducibly deliver the minimum accurate information needed by a voter to assess and maintain active status.
Election Justice USA gathered voter complaints by way of two distinct questionnaires, the differences of which shaped the responses we received. The purpose of these complaint intake forms was to systematically obtain statements of election irregularities which could be easily and rapidly converted to formal declarations or affidavits for legal action. Thus, every voter who registered a report with us understood that they might eventually serve as witnesses or plaintiffs in lawsuits. Hundreds of them did eventually do so.
The first Election Justice USA complaint intake form (New York dataset) targeted New York voters between April 12 – 22 in the wake of Arizona’s infamous primary with a design that reflected the then-emerging understanding that, as in Arizona, there was increasing evidence of involuntary changes to New York State voter registration information as the Primary neared on April 19.2 New Yorkers reported 716 complaints in that short period, certainly understating the Empire State’s difficulties at a time when 120,000 voters were reportedly purged from a single borough’s registered voter rolls.3
Exhibit I of the Election Justice USA’s amended lawsuit in New York includes these summary facts of the 716 entry Election Justice USA database:
- 97 respondents “clearly misunderstood New York’s registration deadlines”
- 619 respondents who did understand the deadlines represented nearly every New York County
- 401 respondents registered from 2012-2016 and legally should not have been subject to voter roll purging
- 303 registered during the current campaign in either 2015 or 2016 and before the relevant deadlines
- 140 of the 619 were switched, without knowledge or consent, to no longer registered
- 289 of the 619 had been switched, without knowledge or consent, to independent
- 79 of the 619 had been switched, without knowledge or consent, to a different party
- 27 of the 619 were simply unlisted at their polling site even though properly registered and active
The New York database of voter registration complaints included a larger proportion of apparent registration tampering cases, some 25 to 30% of the total. Certainly, however, a significant portion of the New York dataset complaints (14%) were submitted by voters who misunderstood the deadlines for party affiliation change (October 9, 2015) or first-time voter registration (March 25, 2016).
A significant number of these complainants provided affidavits for Campanello, et al. vs. New York State Board of Elections, et al.
The second questionnaire (All States dataset) was designed to accommodate a wider array of possible irregularities encountered by voters in New York and from all other state primary elections.4 There remained only 22 scheduled election contests when the form was first posted on April 14.
States with reports of irregularities that constituted possible voter registration tampering:
AZ, CA, CO, FL, IA, IN, KY, LA, MD, ME, NJ, NM, NY, OH, OR, PA, TX, WI
Among the All States group of complaints taken from 1109 voters, the allowance for a variety of reported types of anomaly beyond registration problems led to a lower frequency of involuntary registration switch complaints lodged in the corpus. On average, almost 18% of these reports appeared to constitute likely tampering cases from 18 states.
Election Justice USA worked with two additional databases, both from Arizona, to arrive at a total of more than 2000 complaints of registration related problems. The first of the additional two databases was collected by the online collective Anonymous.5 It included 151 entries, all of which were reviewed in detail by one of the authors of this report. The second database was collected by one of the original members of Election Justice USA’s steering committee and contributed to the beginnings of the Election Justice USA. That Arizona database from Election Justice USA‘s founding included more than sixty entries, a small handful of which overlapped with entries in the Anonymous database. Anonymous summarized their review of the 151 entries as follows:
- 1 Unknown Party
- 12 GOP
- 139 Dem
- 113 Sanders supporters
- 24 Unknown Preference
- 2 Clinton
While a small number (less than five) of the Anonymous entries were attributed to private emails sent to an address set up for that purpose, these numbers are consistent overall with Election Justice USA‘s review of the two hundred or so Arizona entries for which we could obtain all the information.
Eventually, more than 20,000 voters’ affidavit ballots were not counted in Maricopa County alone.6§I.D of this report discusses three lawsuits that have been filed in Arizona to address registration and poll closings in Arizona.
Table 1 presents a summary of the five states wherein the total number of complaints ran into double digits and, together, comprises more than 90% ofthe All States dataset. Including the signature involuntary registration tampering state of Arizona, these five states voted between March 22 and June 7, 2016. Note that All States New York reports are voters who reported through the second form. The rate at which apparent registration tampering was reported was in the 15% range among these states.
Most of the involuntary registration changes of All States complainants involved Democratic affiliations being switched to independent or to unaffiliated with any party while, to a lesser extent, affiliation switching to another party was reported. Almost half of valid New York data complainants were involuntarily switched to unaffiliated. Among All States complainants, a greater rate of registered voters left off of voter rosters was reported than occurred among the five percent of New York reports.
New York Republican voters complained of registration problems at an almost vanishing rate, less than 20 of the 716 New Yorkers, with apparent registration tampering happening at an even lower rate. All States complainants included many more Republicans and other party affiliants but, again the bulk of involuntary registration changes involved Democrats or, in California,so-called No Party Preference voters.
In California, however, there was a large volume of reports from complainants who had been registered to vote in person but who found that their status had been changed to permanent vote-by-mail when they discovered this on arriving at the polls and they were forced to vote provisionally. Details of these complaints are described in more detail in §I.A.2voters and §III.I.
We estimated that this sort of involuntary registration status change comprised roughly 13% to 20% of all complaints we received from voters to the effect that their party affiliation was changed to another or to no party. This proportion of voter complaints rose to some 30% when totals included voters whose registrations were apparently fine until their names went missing from polling place rosters on election day.
Compare such values to a roughly 10% rate of voter complainants having been unregistered or a 13% rate of complainants who’d simply misunderstood or missed the appropriate date for registering as a voter or switching party affiliations.
In conclusion, voter registration tampering was a widespread phenomenon across states holding closed or semi-closed presidential primaries, affecting would-be Democratic voters almost exclusively. It is likely the primary driving factor responsible for the unprecedented number of provisional and affidavit ballots seen in this primary season’s election contests, alongside illegal voter registration purges, which we cover in the subsequent section.