WikiLeaks: Clinton Campaign Talks Rigging Polls

News: US News
by Christine Niles  •  •  October 25, 2016   

Clinton campaign email advises skewing polls by oversampling Democrat-friendly groups

You are not signed in as a Premium user; we rely on Premium users to support our news reporting. Sign in or Sign up today!

Hillary Clinton's campaign has been caught discussing how to skew the polls in favor of Clinton with fake data. A 2008 email released by WikiLeaks Monday shows members of Clinton's campaign discussing strategies to falsely pump the numbers in favor of Clinton.

The email begins with a note from Tom Matzzie, a Democratic consultant in 2008, writing to George Rakis, campaign manager for a George Soros-funded political action committee, and Andy Meyer, research director at the Atlas Project — a political research and data platform for liberals.

Hey, when can we meet? I also want to get your Atlas folks to recommend oversamples for our polling before we start in February. By market, regions, etc. I want to get this all compiled into one set of recommendations so we can maximize what we get out of our media polling. (emphasis added)

In response, Crystal King, an Atlas Project employee, wrote back attaching a 37-page document with recommendations on how to rig the polls, including oversampling Hispanics and the Native American population.
Hispanics and Native Americans tend to vote Democrat. Heavily weighting a poll with an oversized sample from these ethnic groups (without disclosing the fact) gives skewed results in favor of the Democratic candidate. It's a dishonest, unethical way of polling, and misleads the public.
Other examples from the document include recommending monitoring samples in Florida — a battleground state — to ensure "it is not too old, and that it has enough African American and Hispanic voters to reflect the state."
It also suggests polling in Tampa or Orlando, where Independents "are better persuasion targets than north or south Florida."
For national polling, the Atlas Project recommends a sample of 800 "with potential oversamples in key districts/regions." It also suggests "Benchmark polling in targeted races, with ethnic oversamples as needed" and "Targeting tracking polls in key races, with ethnic oversamples as needed."
Watch the panel discuss manufacturing the numbers in "The Download—Skewing the Polls."


Have a news tip? Submit news to our tip line.

We rely on you to support our news reporting. Please donate today.
By commenting on you acknowledge you have read and agreed to our comment posting guidelines