As many as 5.7 million noncitizens may have voted in the 2008 election, which put Barack Obama in the White House.
The research organization Just Facts, a widely cited, independent think tank led by self-described conservatives and libertarians, revealed its number-crunching in a report on national immigration.
Just Facts President James D. Agresti and his team looked at data from an extensive Harvard/YouGov study that every two years questions a sample size of tens of thousands of voters. Some acknowledge they are noncitizens and are thus ineligible to vote.
Just Facts’ conclusions confront both sides in the illegal voting debate: those who say it happens a lot and those who say the problem nonexistent.
In one camp, there are groundbreaking studies by professors at Old Dominion University in Virginia who attempted to compile scientifically derived illegal voting numbers using the Harvard data, called the Cooperative Congressional Election Study.
On the other side are the professors who conducted the study and contended that “zero” noncitizens of about 18 million adults in the U.S. voted. The liberal mainstream media adopted this position and proclaimed the Old Dominion work was “debunked.”
The ODU professors, who stand by their work in the face of attacks from the left, concluded that in 2008 as few as 38,000 and as many as 2.8 million noncitizens voted.
Mr. Agresti’s analysis of the same polling data settled on much higher numbers. He estimated that as many as 7.9 million noncitizens were illegally registered that year and 594,000 to 5.7 million voted.
These numbers are more in line with the unverified estimates given by President Trump, who said the number of ballots cast by noncitizens was the reason he lost the popular vote to Hillary Clinton.
Last month, the president signed an executive order setting up a commission to try to find on-the-ground truth in illegal voting. Headed by Vice President Mike Pence, the panel also will look at outdated voter lists across the nation with names of dead people and multiple registrants.
For 2012, Just Facts said, 3.2 million to 5.6 million noncitizens were registered to vote and 1.2 million to 3.6 million of them voted.
Mr. Agresti lays out his reasoning in a series of complicated calculations, which he compares to U.S. Census Bureau figures for noncitizen residents. Polls show noncitizens vote overwhelmingly Democratic.
“The details are technical, but the figure I calculated is based on a more conservative margin of sampling error and a methodology that I consider to be more accurate,” Mr. Agresti told The Washington Times.
He believes the Harvard/YouGov researchers based their “zero” claim on two flawed assumptions. First, they assumed that people who said they voted and identified a candidate did not vote unless their names showed up in a database.
“This is illogical, because such databases are unlikely to verify voters who use fraudulent identities, and millions of noncitizens use them,” Mr. Agresti said.