The basic facts regarding Wisconsin’s election security:

Real-life Wisconsin miscounts:

  • Lessons from the 2016 Recount (MIT/Harvard/UW Study)
  • Final Report on 2016 Recount (The Stein report)
    These two reports document the results of Wisconsin’s 2016 Presidential recount, which provided a rare opportunity to determine the county canvasses’ error rates. When county boards of canvass were forced to check their work with a full recount, they corrected vote totals they had previously certified in 64% of the state’s precincts. At least 17,681 (1 in every 170) of the originally certified votes were found to have been miscounted. These reports explain that miscounts are common, widespread, and accepted as normal by election officials.
    Note: The high error rate was not reported in commercial news media because the errors randomly went for and against each of the two major-party candidates and roughly balanced out in the final totals–and because commercial media is interested only in the partisan horse race, not in the quality of our election administration.
  • Racine County miscount (2016)
    The scandal was not that the voting machines failed to count as many as 1 in every 40 votes in some precincts–that particular malfunction was well-known to election officials and could easily have been corrected or handled with a work-around.
    The scandal is that the county officials knowingly certified incorrect results twice, and that state officials took no action to improve counties’ canvass procedures.
  • Stoughton miscount (Dane County, 2014)
    A triple-decker misadventure: Most votes went uncounted due to the county’s mistake in setting the machines up to read the ballots; two pre-election tests failed to notice the problem; and poorly maintained machines then counted dust bunnies instead of votes. The story would be funny, except that none of the problems would have been noticed had the mistake caused votes to be miscounted instead of uncounted. The good news: It also demonstrated how very quickly and easily miscounts can be corrected, if they are detected before certification.
  • Medford miscounts (Taylor County, 2004)
    Another miscount caused by incorrect set-up of the machines, but unlike the Stoughton miscount, local officials did not notice that the vote totals made no sense until a researcher from Michigan pointed that out to the county clerk — five months after certification.

External links: