Tomorrow (Tuesday), there's yet another "make-or-break" set of Primaries for the GOP establishment's preferred front-runner Mitt Romney in both Arizona and his home-state of Michigan. While pre-election polls suggest Romney's the run-away favorite in AZ, he reportedly has only a marginal lead in MI. Should he lose that state tomorrow, his party, and their nomination process, could well descend into complete chaos (even more so than it already is.)
But here's the good news for Election Integrity fans keeping an eye on the race track for tomorrow's horse race: Both Arizona and Michigan allow all voters to vote on paper ballots at the polling place if they wish. So there's actually something to count that reflects the voters' intent.
The bad news: Rather than count those ballots, the states will run them through oft-failed, easily-manipulated optical-scan computer systems and report whatever it is that the computers announce as "the results."
While many Americans mistakenly place faith in these systems, a report in today's New York Daily News illustrates --- yet again --- how foolhardy that is. And this one is a doozy (yes, yet another one)...
According to the paper , a few months ago, experts at NYU's Brennan Center for Justice "detected an alarming pattern" in results at a polling place in the South Bronx. The tallies from the November 2010 election were closely examined and it was noticed that there was a high rate of invalidated votes at Public School 65.
New York was the last state in the nation to move from lever voting machines to new op-scan systems a couple of years ago. The 2010 race was the first full use of the new systems. Yet the state Board of Elections, reportedly, did nothing to investigate the concerns about results in The Bronx, as reported by the new systems made by Election Systems & Software, Inc. (ES&S), the largest supplier of electronic voting systems in the country. The company also happens to have a disastrous, years-long record of failed voting systems, featuring entirely lost votes, misreported tallies and, occasionally, impossible numbers reported on them.
With the Board of Elections failing to take action, the New York Daily News made public records requests to inspect the ballots from PS 65 and here are what they describe as their "disgraceful findings":
In the November general election, the scanner handled 289 ballots and misread votes on 156 of them, a 54% failure rate.
The errors occurred in identifying so-called overvotes. These happen when voters fill in two ballot ovals for different candidates in the same race, darkening one for, say, [then Gubernatorial candidate, Andrew] Cuomo and the other for opponent Carl Paladino.
Presented with the conflicting marks, a scanner alerts the voter with an on-screen message that gives two choices: Correct the ballot or proceed. If the voter proceeds, the machine invalidates the vote in that race. Neither Cuomo nor Paladino would be credited with a vote.
Time after time, looking at photographic images of the ballots that are recorded by the scanners, we found ballots that were perfectly filled out: one vote for Cuomo, one vote for Eric Schneiderman, then running for attorney general, one vote for Kirsten Gillibrand, running for Senate.
And, time after time, we also saw that the machine had registered overvotes where none existed. For example, detecting a valid Cuomo vote while also recording phantom votes for Paladino and for the five other lesser-known candidates, plus a write-in.
In those circumstances, the machine invalidated proper votes.
On other ballots, voters chose not to fill out an oval in a particular race — and the machine recorded that they had filled one in. For example, a voter opted to skip the contest between Gillibrand and Democratic primary challenger Gail Goode — but the scanner scored a vote for Goode.
Unfortunately, New York is just now finding out about what we've been warning for so many years concerning computerized optical-scan systems. The state seems disinterested in doing a damn thing about it, if the report from the Daily News is to be believed.
Michigan uses ES&S optical-scanners across dozens of counties (and op-scanners made by different, similarly failed companies, in others). Arizona uses ES&S op-scanners in just two counties and easily-manipulated, oft-failed op-scanners made by other companies in the other counties.
Last December, in a first for the U.S. Elections Assistance Commission --- the federal agency created in the wake of the 2000 Presidential election debacle and charged with testing, certifying and overseeing electronic voting systems in the country --- a warning was released about ES&S op-scanners --- the DS200 systems, the same ones used in NY --- after a study was carried out in the wake of Cuyahoga County (Cleveland), OH, finding that some 10% of their machines failed pre-election testing.
The EAC study found, among other things, in their first investigation of a system they had previously "certified," that the machines would freeze randomly, preventing ballots from being fed into them; machine errors were not recorded in the log files; and that ballots, in many cases, were not read (counted) properly at all.
Apparently the EAC testers hadn't noticed any of those things when supposedly running the system through its paces for federal "certification" testing. Despite the failures they discovered late last year, the disastrous EAC did little more than issue a warning. They didn't bother to decertify the machines that they claimed they had previously tested and "certified" as meeting federal standards. Those same systems, and ones like it, will once again be used across the country in the 2012 Presidential Election cycle.
But those machines, used in OH and NY, were ES&S' newer DS200 systems. In Michigan, much of the state is still using their older M-100 scanners, and will do so again tomorrow.
The BRAD BLOG wrote about the M-100s back in 2008, when Oakland County, MI, was trying to get help from the EAC concerning their precinct-based M-100 op-scanners which, during pre-election testing, had "yielded different results each time" the "same ballots were run through the same machines."
According to the letter [PDF] the county sent to the EAC at the time...
...
ES&S determined that the primary issue was dust and debris build-up on the sensors inside the M-100.
...
Unfortunately, [local clerks] are prohibited from performing any maintenance/cleaning on the machines as it voids the warranties. ES&S has not performed any preventative maintenance under the state contract, since the machines were delivered three years ago.
That was 2008, so it may be seven years since those machines have been cleaned or checked. Back in New York, in 2012, the Daily News notes, with snarky prose that could have been ripped from The BRAD BLOG:
That’s not what happened — and the failures occurred twice. There must be a complete investigation by an independent authority that examines the faulty machine and goes far into checking on the possibility of broader undetected failures.
Yeah, good luck with that.
And good luck to MI and AZ voters tomorrow! Your "results" will likely be ready just as soon as the polls close! And they will either be accurate or inaccurate! Unfortunately, you may not know one way or another for months, years, or ever. The following week we'll be on to Super Tuesday, and it won't really matter if it's later found that hundreds, or thousands of votes were mis-reported by the op-scanners in either MI or AZ --- not that any election officials are even likely to check the hand-marked paper ballots to find out.
But other than that, the U.S. still remains the greatest democracy this planet has ever known.
[Hat-tip to @JeannieDean via the Twitters.]