Home

Wisconsin relies on trust and luck–not verification–to protect its vote totals

by Karen McKim, April 11, 2021

Everyone living in the 21st Century knows computers can malfunction or be misprogammed. We know that problems can arise unexpectedly at any time, and so when computers are used for important tasks, managers need have a reliable way to catch any computer errors. Further, most of us know that voting systems in the 21st century are computers, and that elections technology has no special magic that sets it apart from the same risks that threaten all other computers.

Sensible people expect voting machines, on occasion, to produce incorrect results. It would be weird and unbelievable if they did not. It was no scandal, then, when the cities of Stoughton, Wisconsin and North Kingstown, Rhode Island experienced nearly identical voting-machine mishaps in 2014 and 2016, respectively.

In each city, a municipal yes/no referendum was on the ballot.

Each city used DS200 voting machines manufactured by ES&S. In each city, the computers were accidentally misprogrammed to look for votes in the wrong place on the paper ballots.

In each city, two pre-election tests were performed (by the city and county in Wisconsin and by the city and vendor in Rhode Island). In each city, those tests failed to detect the problem before the computers counted the votes.

In each city, Election-Night results showed the referenda failing, but the machines counted so few votes that the results were literally unbelievable. In each city, municipal officials conducted a hand count of the paper ballots and certified the correct results, so no lasting harm was done. In each city, if the error had been a simple yes/no flip instead of ignoring votes, routine canvass procedures wouldn’t have noticed the problem and officials would have unknowingly certified the wrong outcome.

And that’s where events diverge. Rhode Island news media covered the incident, alerting the state’s civic community to the old adage: To err is human but to really foul things up requires a computer. In response, Common Cause of Rhode Island, the League of Women Voters of Rhode Island, the Rhode Island Board of Elections, Secretary of State Nellie Gorbea, and the Rhode Island Town and City Clerks Association came together to promote state legislation to add what are known as “risk-limiting audits” to that state’s routine canvass procedures.  These efficient audits are not full recounts, but they do ensure that any miscounts serious enough to identify the wrong winner will be detected and corrected before certification. The law passed, and now Rhode Island’s final election results are protected from both accidental and deliberate electronic miscounts. Rhode Island’s election officials can routinely prove that they are certifying only the right winners.

In Wisconsin, however, the Stoughton mishap evoked only one big civic yawn. Only one local newspaper reported the problem, in only a small item on an inside page. Few local officials were even aware of the problem, and those who were aware did not understand what had happened. Civic organizations other than this one took no notice.

Wisconsin’s civic community was willing to accept a risk that Rhode Island’s considered intolerable. 

For nine years I’ve been working on Wisconsin’s election security: study, observation, interviews, and attending national conferences. I can explain a lot. What I cannot explain is why good and sober Wisconsin citizens stubbornly choose to rely on luck and trust to protect our elections, particularly after the events surrounding the November 2020 election demonstrated how very fragile that trust is, and how corrosive disputes get when it breaks. I do not know why Wisconsin election officials still cling to canvass procedures that pre-date computerized tabulation while other states adopt modern canvass procedures, capable of identifying and correcting any serious miscount, and of disproving allegations of incorrect results. I do not understand why Wisconsin voters are satisfied with occasional voting-machine spot-checks that our officials undertake with no intention confirming outcomes and no procedures for correcting any miscounts they find.

I do know that our local election officials are naïve about the risks. Among the quotes I’ve collected: “If these machines were capable of miscounting, the State wouldn’t let us use them,” and “Unless someone has figured out a way to hack through the unit’s power cord, our equipment is basically unhack-able.”

Wisconsin election officials are demonstrably naive about national developments in election auditing. In September 2018, the Wisconsin County Clerks Association adopted a resolution that displayed ignorance of basic election-security recommendations such as those from the Presidential Commission on Elections Administration and the National Academies of Sciences (2018). One of my most jarring moments came when a Wisconsin municipal clerk argued with me by saying: “Yes, the convenience store reconciles its cash register every night, but they are counting money, not just votes.”

Why do clerks in other states stay abreast of advances in election security while Wisconsin’s clerks do not? Where is Wisconsin reporters’ skepticism, or even curiosity, when those clerks say things like “the equipment is never connected to the internet,” even as the reporters see Election-Night results coming out of the county computer moments after being tabulated by the polling-place computers? Why do civic organizations in other states promote effective election security, while their counterparts in Wisconsin promote mere trust? I truly do not know.

When I got involved in 2012, I had hoped to help make Wisconsin among the first in the nation to secure its elections with effective canvass procedures. Now, watching the taillights of states like Georgia, Maryland, and Florida modernize their canvass practices and leave Wisconsin’s in the dust, I can only hope Wisconsin is not among the last.

And that is the reason for placing this personal note on the home page of the Wisconsin Election Integrity website. My husband and I are planning to leave the state before the close of 2021, and I am winding down my activities. I am transitioning this website from an active status to one that will, I hope, be useful to others who might someday take up the banner, and:

1) contribute a vocabulary and framework to an informed and productive civic discussion that will advance the cause of election-securing canvass procedures in Wisconsin;

2) enable and motivate Wisconsin media to cover the story as a public service, even though the issue has no inherent partisan frisson; and

3) enable and motivate civic-minded individuals and organizations to pick up the cause after I leave.


Explore the links in the menu above for more information.

If anyone is interested in purchasing this url —and in using it for a genuinely nonpartisan effort to promote practices endorsed by nonpartisan election-security authorities— please contact me at mail.k.mckim@gmail.com.



*  *  *