Crosscheck's Flawed Data as "Proof"
Election officials who believe double voting is a great risk to electoral integrity have had access to Crosscheck for a decade but have delivered a very small number of prosecutions and convictions of double voting.
This small number of convictions, despite commitment of massive time and resources, suggests that Crosscheck is ineffective at identifying double voting.
Yet, political pundits and lawmakers continue to make broad claims about double voting as a substantial risk; in circular fashion, they often cite Crosscheck's flawed data as proof of or risk from widespread double voting.
A few examples from "credible" sources are listed below which illustrate the danger done by the overly broad match methodology employed by Crosscheck. (For a sense of the scope of data abuse, scan less credible media to see what they made of the North Carolina numbers.)
In 2015, Kansas Secretary of State Kris Kobach claimed that Kansas has "125,000 voters also registered in another state". (8:40 mark).
125,000 was the number of "Potential Double Registrants" pairs reported by Crosscheck for Kansas in 2015.
We estimate that 60%-90% of those pairs are two unique voters misidentified by Crosscheck.
Conservatively, the 125,000 pairs reported might represent 50,000 doubly registered Kansas, plus 75,000 Kansans falsely matched with 75,000 American voters from another Crosscheck state despite mismatched or missing social security number or other identifiers.
National Review published the headline "N.C. State Board Finds More than 35K Incidents of ‘Double Voting’ in 2012".
The 35,000 records reported by North Carolina State Board of Elections represented voters matched solely on first and last name and DOB. Statistically, they are almost certainly 70,000 individual voters who share a name and DOB.
Four individuals were ultimately referred for prosecution for double voting.