Quantifying Illegal Votes Cast by Non-Citizens in the Battleground States of the 2020 Presidential Election
By James D. Agresti
November 8, 2020
Based on current population data from the Census Bureau and voting data from previous elections, Just Facts has conducted a study to estimate the number of votes illegally cast by non-citizens in the battleground states of the 2020 election. The results—documented in this spreadsheet—show that such fraudulent activities netted Joe Biden the following extra votes in these tightly contested states:
- Arizona: 51,081 ± 17,689
- Georgia: 54,950 ± 19,025
- Michigan: 22,585 ± 7,842
- Nevada: 22,021 ± 7,717
- North Carolina: 46,218 ± 16,001
- Pennsylvania: 32,706 ± 11,332
- Wisconsin: 5,010 ± 1,774
If the lower end of these illegal vote estimates were removed from the vote tallies as of November 8, 2020, 2:00 AM EST, Donald Trump would be leading in states that have a total of 259 electoral votes, or 11 shy of the 270 needed to win the presidency. If the upper end of the illegal vote estimates were removed, Trump would be leading in states that have 285 electoral votes, or 15 more than needed to win the presidency.
These estimates account for just one type of election fraud, and they tend to understate it because they depend on Census surveys, which are known to undercount non-citizens.
Just Facts asked two Ph.D. scholars who specialize in data analytics to critically review this study, and they assessed it as follows:
“I find this research of great value—clear in its assumptions, clear about the sources of data used, methodologically sound, and fair in its conclusions. Furthermore, it contains enough references to allow any interested person to ‘fact-check’ every aspect of it.”
– Michael Cook, Ph.D. Mathematician, Scientific and Quantitative Researcher
“Instead of adding politics, vitriol, and bias to this timely, heated topic, this study provides a credible data analysis that supports a strong hypothesis of non-citizens having a significant effect on this election. Any serious critic should try improving on these estimates, as opposed to dismissing them with unproven claims.”
– Dr. Andrew Glen, Ph.D., Professor Emeritus of Operations Research, The United States Military Academy, and Award-Winning Researcher in the Field of Computational Probability
Pathways to Illegal Voting
All 50 states require people to be U.S. citizens in order to register to vote in federal elections, and federal law forbids non-citizens from falsely claiming U.S. citizenship to register to vote. However, enforcement mechanisms for such laws are limited, and opportunities to get around them are ample.
The federal voter registration form requires people to declare under penalty of perjury that they are U.S. citizens, but it does not require them to provide documentary evidence that they are citizens. Several states, including Arizona and Georgia, tried to implement such a requirement, but they were blocked from doing so by court rulings backed by the Obama administration.
The enforcement situation was aptly summarized by Barack Obama shortly before the 2016 U.S. presidential election when actress Gina Rodriguez asked him if “Dreamers” and “undocumented citizens” would be deported if they voted. Obama replied:
Not true. And the reason is, first of all, when you vote, you are a citizen yourself. And there is not a situation where the voting rolls somehow are transferred over, and people start investigating, etcetera.
After dodging the fact that Dreamers and undocumented immigrants are not citizens, Obama’s clear message was that there is no effective way to enforce the law that prohibits them from voting.
Some states require applicants to submit their Social Security numbers, but the vast majority of states allow for other means of voter registration, like submitting a driver’s license number, bank statement, or utility bill. These provide little-to-no evidence of citizenship.
Even if a state requires a Social Security number and actually checks it, this does not constitute clear proof of citizenship because many non-citizens have illegally obtained Social Security numbers. In 2013, the chief actuary of the U.S. Social Security Administration estimated that in 2010:
- 0.7 million illegal immigrants worked by using Social Security numbers obtained by using “fraudulent birth certificates.”
- 1.8 million illegal immigrants worked by using Social Security numbers “that did not match their name.”
The magnitude of identity fraud that would allow non-citizens to vote has been laid bare by:
- a 2002 U.S. Government Accountability Office investigation that found:
- “the use of fraudulent documents by aliens is extensive.”
- in November 1998, immigration officials in Los Angeles “seized nearly two million counterfeit documents, such as … permanent resident cards and Social Security cards, which were headed for distribution points around the country.”
- a 2005 New York Times article that reported: “Currently available for about $150 on street corners in just about any immigrant neighborhood in California, a typical fake ID package includes a green card and a Social Security card.”
- the following statement made in 2017 by California Senate Leader and Democrat Kevin De Leon when testifying before the Senate’s Public Safety Committee:
I can tell you half of my family would be eligible for deportation under [Trump’s] executive order, because if they got a false Social Security card, if they got a false identification, if they got a false driver’s license … if they got a false green card. And anyone who has family members who are undocumented knows that almost entirely everybody has secured some sort of false identification.
Illegal Voting Data
In accord with the open doors to illegal voting, scientific surveys of non-citizens have found that:
- 15% admitted they were registered to vote in 2008, and 8% stated “I definitely voted” in the 2008 U.S. presidential election.
- 14% admitted they were registered to vote in 2012, and 9% stated “I definitely voted” in the 2012 U.S. presidential election.
- 13% of Hispanic non-citizens admitted they were registered to vote in 2013.
The consistency of these rates over time suggest that they are applicable to later elections where such data is currently unavailable.
Furthermore, those rates are merely for self-admitted actions, and database matches with voting and registration records show the actual rates are about twice as high. In 2008, the one year for which Just Facts has full data, 27% of non-citizens were registered to vote, and 16% of them actually voted. This 16% rate, applied to the latest Census data for the numbers of non-citizens in battleground states, is the basis for the 2020 estimates of illegal voting by non-citizens.
The 16% rate is partially corroborated by the 2012 data. In that year, 9% of non-citizens stated “I definitely voted” in the presidential election, and database matches could add another 11% to this figure. However, some of the people in the database overlap with those who said they voted, and the exact number of overlaps is unknown to Just Facts. If the overlap rate is the same as in 2008, the data would show that 17% of non-citizens voted illegally in the 2012 election.
The results of such studies have sizeable margins of error due to relatively small sample sizes, and these are reflected in the above estimates for the 2020 election. As is often the case with estimates of illicit actions where enforcement is sparse, there are other possible sources of error, some of which may produce overcounts and some undercounts. But given that the Census Bureau counted 21,749,984 non-citizens in the U.S. during 2019, even if only 5% of them voted, this would amount to more than a million illegal votes cast in the 2020 election.
Some media outlets and “fact checkers” have tried to contest the reality that substantial numbers of non-citizens vote in U.S. elections, but a multitude of facts from academic books and journals have shown that their arguments consist of mathematically illiterate notions, half-truths, and outright falsehoods.
Who Do Non-Citizens Vote For?
In 2008, 82% of non-citizens who said they voted also stated that they voted for Democrat Barack Obama, and 18% said they voted for Republican John McCain.
Those votes are consistent with the Democratic Party’s push to expand social programs and a 2011 scientific survey of Hispanic immigrants which found that 81% of them want “a bigger government providing more services,” and 12% would prefer “a smaller government with fewer services.” In contrast, 41% of the general U.S. population want a bigger government, and 48% would prefer a smaller one.
The lopsided votes of non-citizens for Democrats are also consistent with the promises and actions of Democratic politicians to give free healthcare, amnesty, and citizenship to people who illegally immigrate to the United States.
Just Facts’ study applies the party vote breakdown of 82% Democrat and 18% Republican to the 2020 election. By subtracting these illegal votes for Trump from the illegal votes for Biden, the study arrives at the figures above for the extra votes that these fraudulent activities have netted Biden in the battleground states.
Trump’s Election Integrity Commission
A common argument used to dismiss facts about election fraud is that President Trump’s Advisory Commission on Election Integrity failed to find widespread evidence of such malfeasance. This claim is a classic half-truth because it neglects to reveal that the Commission existed for less than a year because its work was blocked by the refusal of states to turn over voter data and a flurry of lawsuits.
In the words of California’s Secretary of State in 2017:
While the commission is allowed to request the personal data of California voters, they cannot compel me to provide it. Let me reassure California voters: I will not provide the Commission with any personal voter data. …
Yesterday’s ruling is merely the first in a string of lawsuits challenging the Commission. Those lawsuits send a strong message—the Commission will face opposition at every step of the way from those who are fighting to protect our voting rights, our privacy, and our democratic principles.
Contrary to the allegation that the Commission sought “personal voter data,” it actually asked the states for “detailed, publicly available voter roll data.” Given that Trump now has legal “standing,” or “a personal stake in the outcome of the controversy,” courts may force states to turn over the data that they previously withheld. This would allow the Trump campaign to cross check voter rolls against other databases that contain information on citizenship status.
Without such cross checks—which certain states stridently opposed under the guise that the data is personal—claims that widespread voter fraud doesn’t exist because the Commission didn’t find it are misleading and baseless.
A wealth of facts show that there are ample openings for non-citizens to illegally vote and that roughly 16% of them voted in the 2008 and 2012 presidential elections. If this was also the case in the 2020 battleground states where Biden currently has slim leads, the numbers of non-citizens in these states and their preference for Democrats may have netted Biden enough fraudulent votes to tip the overall election winner from Trump to Biden.
Illegals, plus the sale and harvesting of legitimate ballots is enough to add 1-3% in Blue states. In person, one day, paper ballots, absentee with cause, purple fingertip is the cure. PS End voting machines!
All Americans should be outraged at what is going on with this election. At the very foundation of our system is the premise that we have a honest, free, open and transparent election. What happened this time is nothing more than corruption on grand scale. No wonder we have such strife in this country. Our nation will not stand if this is allowed to go on. Don’t know when, but we will crash and burn if it’s not fixed.
GOP should be organizing street protests like the DNC did many times.
Where are our representatives?!
I AM outraged. I’m outraged we have a president and sycophantic followers who will simply say anything to undermine the integrity of our democratic process.
Put up or shut up.
This “study” is not a study. It is an opinion supported by a skewed, bias regression of data based on dubious assumptions. And you know what it means to make a false assumption. An assumption that even if there were illegal votes, they would have all gone to Biden is not substantiated. Trump did better among minorities than he did last time.
The article proves that your comments are false and explains exactly why they are false. Read it carefully. I’m not going to spoon feed it to you.
i am sorry, but i did read it very carefully. i did not see any proof of your claim i wanted too. A wealth of facts show that there are (ample openings for non-citizens to illegally vote) is not proof that they did or who they would vote for.
Your comment reveals that you didn’t understand the article or are deliberately misrepresenting it. See the sections labeled “Illegal Voting Data” and “Who Do Non-Citizens Vote For?”
Don’t waste your time, Mr Agresti, this person is obviously a dolt who does not understand basic logic. But it’s not his fault — logic is no longer taught in schools; indeed the opposite is taught… Now sing the unbearable 1975 pop hit “Feelings…”
“In 2008, 82% of non-citizens who said they voted also stated that they voted for Democrat Barack Obama, and 18% said they voted for Republican John McCain.”
“A wealth of FACTS show that there are ample openings for non-citizens to illegally vote and that roughly 16% of them voted in the 2008 and 2012 presidential elections. If this was also the case in the 2020 battleground states where Biden currently has slim leads, the numbers of non-citizens in these states and their preference for Democrats MAY have netted Biden enough fraudulent votes to tip the overall election winner from Trump to Biden.”
Your patently false statement that the study implies that ALL non-citizen votes would go to biden makes your “outrage” ridiculous and is based on your own lie. And the fact that Trump did better with legal minority voters isn’t even applicable as the things that Trump did for the legal citizens of our community does not apply to non-citizens.
Reading comprehension is a beautiful thing if you can try and keep up
Had you bothered to read it instead of just whine about it, you would see that the study takes into account that not all votes cast by illegal aliens would have gone to Trump.
This should be investigated completely by all agencies so deemed as the authority to do so.
Furthermore, it must be published by all social media inclusive of national TV, internet and newspapers until all results of the election have been audited to the satisfaction of all interested and connected parties.
Is this article from the same Jim Agresti I recognize?
Thank you for recognizing the import of this study. I don’t think we know each other, but here is my bio: https://www.justfacts.com/james.d.agresti.asp
Has it occurred to anyone that since the census didn’t get completed, all of your suppositions regarding numbers of legal vs illegal voters are purely speculation?
Why didn’t it get completed?
If you took the time to carefully read the article, you would have seen that it addresses this very point. Moreover, it strengthens the finding of the study. Per the 4th paragraph:
“These estimates account for just one type of election fraud, and they tend to understate it because they depend on Census surveys, which are known to undercount non-citizens.”
Good work James
Please do a cross check on all data to confirm your findings, this needs to solid!
Thank You for your hard work
Thank you for your kind words. We don’t publish anything without cross checking anything that can be cross checked.
IF JOE BIDEN WON THE PRESDENCY
Say I flipped a coin 1,000 times for Trump or Biden and 97% came up for Biden. Also you can’t watch me flip the coin even though the law says you can; you can’t. Just trust me, he got 97%. (now shut up and home)
Looking at the math used for financial crimes or regulating any large number sets their was massive cheating or fraud. Proving it traditionally after the fact when you have been blocked from watching the process is hard, but beating the numbers by 47% is impossible. We have president elect thief Biden, for now.
The Democrat party has “normalized” election fraud. For years, there have been illegal votes cast (always benefiting Democrat candidates), but it is believed to be “not material” and the courts wouldn’t overturn a result. But now we see something entirely different. The Dems have implemented massive and wide-spread systems of fraud. They have used this virus to steal a presidential election with massive mail-in fraud, and they may have used intelligence tools (Hammer and Scorecard) to digitally alter votes. “Any” election fraud is treason. There should be mandatory life sentences imposed for any conviction. It will destroy our country if it is allowed to continue, and if they are successful stealing this one, we may never be able to stop them.
Great work Just Facts. It is curious that we aren’t hearing about the blockchain encryption codes in the watermarks on ballots. CISA – Cybersecurity And Infrastructure Security Agency. The Hill reported on 11/21/19 that DHS has invested in an election auditing tool to secure the 2020 elections.
I have been thinking something was up since we haven’t heard a peep from Bill Barr. He was red in the face furious months ago when Wolf repeated the “mail in ballots are safe” absurdity and now we are to believe he could care less about it? I don’t think so. I bet they are letting it play out so treasonous media can hang themselves.
Why is media demanding we show them evidence? Isn’t digging up the facts their job?
We The People own the vote not The Cheaters and Their Woke!
I suspect that the reason that we are not hearing about watermarks and block-chain is that it is a conspiracy theory…
My two mail in ballot that I didn’t use had watermarks on them. You can find them with a black light.
they need to throw out the entire election and start over. this time do it with full transparency. put the military in if need be to make sure it is done fairly and properly.
at the very least, do a recount in every contested state.
Michigan passed same day registration and a curious change to penalties for lying about legal status- see Michigan SB 1238, Pg. 8 warning on citizenship status- is now a misdemeanor penalty in 2018, but federal law 18 U.S. Code Section 611 3(b) says any person who violates this section shall be subject to a fine or one year in prison, or both. Those who attempt to vote but are not citizens are inadmissible and deportable.
*Washington passed their Access to Democracy package in 2018 also. You might begin to believe the Democrat Party was coordinating these changes in voting? It provides for automatic registration when getting an enhanced driver’s license and same day registration. Who can check citizenship status the same day in either state? In the bill covering automatic registration ( for enhanced drivers licenses that act as a passport level of citizenship) HB 2595 Section 206 (1) (2) RCW 29a.08.370 changes the penalty for willfully and knowingly lying about legal status and provides for no penalty at all other than notifying the secretary of state to remove the voter from the rolls. Both of these laws are unconstitutional because no state may write laws that supplant federal law.
Your research is excellent but there isn’t enough time to do the research with our courts divided by who nominated the judges. The laws above are illegal in their present form. We are heading to a point, where doing things legally as should be, are to time consuming. Joe Biden is a documented criminal, so are Hillary Clinton’s actions. Violence may be the only solution as it was when our forefathers wrenched this country out of the grip of the King of England.
Vote early & Vote often ………… the democrat creed
The spreadsheet shows that this was based on the same 2014 study which was used to claim that millions of people voted illegally for Hillary Clinton. That use of the study was debunked as shown in Snopes.com
see https://www.snopes.com/fact-check/hillary-clinton-800000-votes-non-citizens It is shameful that this site is using already debunked statistics to cast doubt on the validity of the electoral process.
As the article documents: Some media outlets and “fact checkers” have tried to contest the reality that substantial numbers of non-citizens vote in U.S. elections, but a multitude of facts from academic books and journals have shown that their arguments consist of mathematically illiterate notions, half-truths, and outright falsehoods.
Thus, you may want to consider the possibility that your view of this article is the result of others misleading you. As Just Facts has documented, Snopes, FactCheck.org, PolitiFact, and the Washington Post Fact Checker have repeatedly misinformed their readers.
Fortunately, the most important factor in making these reported statistics not valid comes from the raw data. Consequently, understanding the problem does not require sophisticated understanding of statistics. In the raw data of 19,000 respondents a total of 121 identified themselves as immigrant non-citizens. By itself, that sample is too small to derive any meaningful statistics that could be used to predict the percentage of non-citizens who voted in ANY election, much less state by state ten years later. The biggest problem, however, is that the responses to questions about citizenship indicate that the respondents did not understand the questions and therefore answered inconsistently. The study was longitudinal, meaning that the questioners were able to match the individual respondents in 2010 with their 2012 responses. Of the people who identified themselves as immigrant non-citizens in 2012, 20 had answered in 2010 that they were citizens. That would have been almost impossible as very few people lose their US citizenship. Your entire article is based on the 14% of the 120 or so respondents who said that they were non-citizens and also that they were registered to vote, but least 16.5% of the respondents gave wrong answers on the citizenship question. If their second answers were correct, they SHOULD have ben registered to vote. Taking the 14% from the tiny sample where the inconsistencies in data are greater than 14% and then using it to say how many people voted illegally in 2020 is such an obvious abuse of statistics that it should be considered fraudulent.
Your arguments are debunked at this link I provided above and here as well. Read carefully. I’m not going to spoon feed this to you.
James, First, I thank you for showing my comments and responding. I read the material in your two links. They do not “debunk” my arguments. First there is a basic problem with your use of statistics. You are basing your claims on a sample which, for 2008, had 67 people who self-identified as non-citizens and either self-reported as being registered or had their registration status verified and of whom only 38 said that they voted. You have an unstated assumption that the data for 2008 (which was not separated by state) is homogeneous with the data for each state in 2020. That assumption is very unlikely to be valid.
Of the 38 respondents who reported that they voted, registrations as voters could only be found for 5. The excuse that you provide for the difference is that many illegal aliens have illegally obtained identification documents. Immigrant non-citizens, however, include legal residents such as Green Card holders. In order for your excuse to support your estimates of illegal votes, you would have to assume that the non-citizens, including a proportionate number of green card holders, registered to vote under fictitious names. I do not find that plausible. Why would a non-citizen want to take that risk just to perform a civic duty? Why would Green Card holders take that risk? If it is easy for people to register under fictitious names and they want to illegally vote, then why don’t they (and citizens besides) do that under multiple names and vote many times?
The inconsistency in the raw data itself remains as my most important argument. That data was reported as having a significant number (20 in 2010) of the people who said that they were non-citizens in a later year having reported that they were citizens in an earlier year. You never really address this other than to say that Brian Schaffner’s estimate of zero voters among immigrant non-citizens is not valid. I agree that the estimate of zero is not valid, but I would say that the data in this study does not invalidate a null hypothesis that the number is zero. The respondents who clearly erred in their survey responses by reporting that they were citizens in 2010 but not in 2012 are unlikely to be homogeneous with the rest of the sample. They may have all been voters for years when they errantly reported that they were non-citizens. If so, they would constitute the majority of the respondents who identified as immigrant non-citizens but also reported that they voted. Furthermore, the fact that 20 respondents obviously erred in reporting their citizenship leads me to suspect that many other respondents made that error without the inconsistency which makes the error obvious.
In conclusion, taking the 38 respondents who said that they were immigrant non-citizens and voted, and then extrapolating (as you are doing) to apply that state by state in 2020 while knowing that there are inconsistencies with the data leads to conclusions which are not supported by the data.
You either failed to understand the articles or are deliberately misrepresenting them. They do, in fact, explicitly debunk your arguments except for this new one you just added, which is blatantly false: “You have an unstated assumption that the data for 2008 (which was not separated by state) is homogeneous with the data for each state in 2020.”
That’s not an “unstated assumption” but a conclusion backed by data that is clearly presented in the article above:
And the very last paragraph of the article states:
These are among the reasons why the article calls the figures “estimates” five separate times.
Unstated assumption? Surely, you jest.
I urge readers to carefully read the article in the 2014 article in Electoral Studies from which you make your extrapolations to 2020, the subsequent reports on the raw data that was used for the articles, my comments, and your responses. Then, the readers who have good mathematical aptitudes should be able to make their own decisions as to whether your article or my comments constitute bunk.
Sorry but your basic assumption that only 18% of non-citizens (primarily Hispanic) would have voted for Biden is severely flawed. Exit polls showed that over 40% of Hispanics voted for Trump in 2020, up from 28% in 2016. Using a previous election’s demographics (Obama vs McCain was 8 years ago) as the backbone for your numbers was in my opinion, a cherry picked statistic that needs to be explained before it can be trusted.
What’s severely flawed here is your assumption that Hispanics who are not citizens vote like those who are citizens. Non-citizens are far poorer, far less educated, and much more likely to favor amnesty and more welfare. Biden embraced such polices, while Trump strongly opposed them. Given these facts, it is absurd to assume that increased Latino support for Trump came from non-citizens.
As a small-l libertarian I actually agree with some of your policy analysis in terms of their necessity, albeit usually for different rationales. However, as someone who practiced immigration and criminal defense I believe that your analysis seems to be based on the view of an immigration and legal system that exists at best on paper only and in reality, due to the nature of administrative law and custom and congressional abdication, without being a practioner in the field whose job it is to keep up with these, the analysis you posted in this particular sector are flawed in ways that start fundamentally with the problem that what in the immigraiton context words do not mean what words are defined as in dictionaries, and same goes for the criminal context. I don’t know if you are at all interested in actually getting an accurate portrayal as to the ambiguities and the complications involved in even proving citizenship status itself, not to mention everything else, but it doesn’t feel like something that is short enough to fit in a comment box. Without debating as to whether the overall policy is good or bad, and without seeing your email address listed here, feel free to drop me an email and I’d be happy to shed some light as to the doublespeak that seems to have completely fouled the datasets in question on the definitional aspect alone. So yeah, you have my email here, let me know if you want a comprehensive definitional explainer on what’s likely the most kafkaesque field of law in American law and regulations, possibly on par with tax law. Shoot me an email if you’re interested. Cheers.
Thank you, but I’ve scrutinized the definitional aspects of immigration and have authored extensive research on this topic. Make your point here, and if it is has merit, I’d be happy to respond.