Darren Hauck via Getty Images
It turns out that there are a few new reasons for Wisconsin to proceed with their recount, and they have nothing to do with Jill Stein’s recount petition. Yes, she filed the petition and yes, she paid the fee, but the state needed to do this even without Stein’s request. A new analysis in the Washington Post discovered significant voting anomalies in several wards.
Trump now leads Hillary by just over 22,000 votes in Wisconsin. The WaPo analysis didn’t look at precincts or counties – they looked at wards, which are the smallest unit where votes are counted. This helped create a more detailed picture of what’s happening there.
Walter Mebane, an associate researcher at the Center for Political Studies and a professor of political science and statistics at the University of Michigan, looked at small wards that used optical scanning technology, and found that certain features in the vote tabulations reveal the possibility that these counts were tampered with.
For instance, using a method called “last digit diagnostics,” Mebane found that the tabulations he looked at should have last digits that, when averaged together, show a mean of around 4.5. Lower could mean that the counts were manipulated.
In the small wards he looked at where optical scanning tech was used, the last digit diagnostics reveal a mean average much lower than 4.5.
Another statistic Mebane looked at was how often the last digit of a vote tabulation was zero, or five. If there are no problems, then the mean average of that variable should be around 0.2. Larger could mean someone was sloppy, and smaller could mean votes were manipulated.
In these small wards, that number was significantly smaller than .2 for Hillary, meaning that the vote counts show a last digit of zero or five far less often than they should.
He also found evidence of something called “signaling” in the tabulations, which is when fraudsters leave a type of trail that more or less claims credit for the fraud. Mebane points out that this is actually a fairly common tactic amongst fraudsters in Russia.
Finally, a simple test revealed something called “multimodality,” which could show that someone was receiving fraudulent votes. These small wards exhibited that anomaly, too.
None of this is absolute proof of election fraud or vote fraud in Wisconsin, and Mebane is careful to note that. However, it gives credence to the need for a recount, whatever Jill Stein’s motivations are. If nothing else, conducting a thorough recount will lend further legitimacy to our democratic process or concretely reveal problems that need to be fixed. It might also serve to weaken Trump’s efforts to undermine our democracy by repeatedly crying “FRAUD!! FRAUD EVERYWHERE!!”
The recount begins today.