Well this is about as sur­pris­ing as a sun­rise in the morn­ing:

By Brad Fried­man on 12/23/2011 2:23pm
U.S. EAC Finds ES&S Paper Bal­lot Scan­ners Used Across the Coun­try Fail to Count Votes Cor­rect­ly
For first time, com­pro­mised fed­er­al agency admits their pre­vi­ous cer­ti­fi­ca­tion was deeply flawed...

Not to say we told ya so, but, ya know, we’ve been telling you so for years (and years.)

A new find­ing by the U.S. Elec­tion Assis­tance Com­mis­sion — a rare find­ing, first of its kind, in fact, as the woe­ful EAC has nev­er before tak­en the time to inves­ti­gate and report on seri­ous fail­ures of e‑voting sys­tems that it has pre­vi­ous­ly blessed with fed­er­al cer­ti­fi­ca­tion — reveals that ES&S paper bal­lot opti­cal-scan sys­tems used in a bunch of large swing states, result in machines freez­ing dur­ing elec­tions, fail­ing to log sys­tem events cor­rect­ly, and, per­haps most trou­bling, bal­lots being mis­read and votes being lost entire­ly.

The EAC’s “For­mal Inves­ti­ga­tion Report” fol­lows on April 2010 rev­e­la­tions by the Cleve­land Plain Deal­er that some 10% of Cuya­hoga Coun­ty (Cleveland)‘s EAC-cer­ti­fied ES&S Precinct Count Opti­cal Scan­ner and Uni­ty 3.2.0.0 tab­u­la­tor vot­ing sys­tems failed in pre-elec­tion test­ing last year.

The paper bal­lot scan­ning com­put­ers were pur­chased as a replace­ment for the 100% unver­i­fi­able Diebold touch-screen sys­tems used pre­vi­ous­ly in Ohio’s largest coun­ty, after a mas­sive analy­sis of all of the state’s e‑voting sys­tems, over­seen by for­mer Sec. of State Jen­nifer Brun­ner (D), revealed seri­ous secu­ri­ty issues and oth­er major flaws in the touch-screen vot­ing machines used there and in many oth­er states.

Relat­ed­ly, Diebold’s own paper bal­lot opti­cal-scan sys­tem has sim­i­lar­ly been found, in the past, to include a flaw which results in votes being lost entire­ly, though the EAC nev­er issued a warn­ing about that sys­tem, to our knowl­edge, even after it led to hun­dreds of votes going uncount­ed in at least one elec­tion in North­ern Cal­i­for­nia (and lord only knows how many else­where that the same sys­tem is used.)

The new find­ings of the fail­ures of the ES&S op-scan sys­tem led Plain Deal­er reporter/blogger Lau­ra John­ston today to wor­ry: “If the com­pa­ny can’t cor­rect the flaw, the gov­ern­ment could decer­ti­fy the machines — leav­ing Cuya­hoga and juris­dic­tions [through­out] the coun­try no way to con­duct elec­tions in a pres­i­den­tial year.”

Um, did the cit­i­zens of Cleve­land lose their eye­balls? Or the abil­i­ty to add 1 + 1 + 1, Ms. John­ston? Yes, there are oth­er ways “to con­duct elec­tions in a pres­i­den­tial year.” For exam­ple, one could sim­ply count the bal­lots by hand in pub­lic, at the precinct, in front of all vot­ers, all par­ties and video cam­eras, and report the results right then and there before the bal­lots are moved any­where — just as they still do in some 40% of the towns in the “First-in-the-Nation Pri­ma­ry” state of New Hamp­shire.

The flawed scan­ners man­u­fac­tured by ES&S, the nation’s largest e‑voting ven­dor, are cur­rent­ly set to be used again in 2012, not only in Ohio, but also in Flori­da, Illi­nois, Indi­ana, Ohio, New York and Wis­con­sin, among oth­ers states...

...

@Pterrafractyl: It looks like peo­ple are real­ly start­ing to wake up. The Dix­ie­crat-ified GOP is going to have a LOT of explain­ing to do one of these days. =)

Ah, the ol’ Abramof crew: the gift that keeps on tak­ing giv­ing:

TPM
Vot­er Fraud ‘Study’ Authored By Repub­li­can Who Plead­ed Guilty In Abramoff Scan­dal

Ryan J. Reil­ly August 3, 2012, 6:13 AM

A new paper claim­ing that vot­er ID laws actu­al pro­tect rather than dis­en­fran­chise minor­i­ty vot­ers is get­ting play in con­ser­v­a­tive cir­cles. What isn’t being men­tion so much is the back­ground of the paper’s author.

Horace Coop­er, the author of the paper, told the Dai­ly Caller this week that vot­er fraud “crim­i­nals — more often than not — are Democ­rats vio­lat­ing the rights of peo­ple who tend to be black or senior.”

Coop­er may not have any exper­tise on vot­er fraud, but he does know a thing or two about fal­si­fy­ing doc­u­ments. Coop­er was indict­ed in 2009 on five pub­lic cor­rup­tion charges, charged with exchang­ing polit­i­cal favors for gifts from Jack Abramoff. Coop­er alleged­ly accept­ed bribes as a staffer to for­mer Major­i­ty Leader Dick Armey, as chief of staff for Voice of Amer­i­ca and when he worked for the Depart­ment of Labor. Coop­er lat­er plead­ed guilty to a mis­de­meanor charge of fal­si­fy­ing a dis­clo­sure report and was sen­tenced to 36 months of pro­ba­tion.

Coop­er asked a judge to short­en his sen­tence ear­li­er this year so he could trav­el to Cen­tral Fiji on a trip paid for by Qorvis Com­mu­ni­ca­tions, writ­ing in a court doc­u­ment that he was hope­ful a “more per­ma­nent employ­ment rela­tion­ship may devel­op” after the trip. A judge denied the motion. Cur­rent­ly, he is “self-employed as a writer and legal com­men­ta­tor, and oper­ates a PR con­sult­ing ser­vice,” accord­ing to a court fil­ing.

In an inter­view on Thurs­day, Coop­er accused TPM of “play­ing gotcha games” when we inquired about whether he was still on pro­ba­tion and said there was “absolute­ly no rel­e­vance” to the issue at hand. He called the Jus­tice Department’s orig­i­nal charges against him “unsub­stan­ti­at­ed” and said he’s been inter­est­ed in vot­er integri­ty for a long time.

...

Umm...Earth to John, the mic’s on:

TPM
John Fund: Sure, Repub­li­cans Focus On Vot­er ID For Polit­i­cal Rea­sons

Ryan J. Reil­ly August 7, 2012, 3:16 PM

John Fund, the for­mer Wall Street Jour­nal colum­nist who has been pro­mot­ing vot­er ID laws for years, admit­ted Tues­day that some Repub­li­cans focus on vot­er ID laws which restrict in-per­son vot­ing over laws which could lim­it absen­tee vot­ing because the GOP has a per­ceived elec­toral advan­tage when it comes to vot­ing by mail.

“Absen­tee vote bal­lot fraud is the tool of choice amongst fraud­sters,” Fund told a group of blog­gers munch­ing on Chick-Fil‑A at the con­ser­v­a­tive Her­itage Foun­da­tion on Tues­day. “Every­time you see a tru­ly mas­sive, coor­di­nat­ed effort at vot­er fraud, it usu­al­ly relies in part on absen­tee vot­er fraud.”

Fund said that many vot­er ID laws “take some pro­vi­sions to curb absen­tee bal­lot fraud,” with a few excep­tions. But he con­fessed that Democ­rats had a point when they say that Repub­li­cans focus on vot­er ID because of a poten­tial elec­toral advan­tage.

“I think it is a fair argu­ment of some lib­er­als that there are some peo­ple who empha­size the vot­er ID part more than the absen­tee bal­lot part because sup­pos­ed­ly Repub­li­cans like absen­tee bal­lots more and they don’t want to restrict that,” Fund said. “But the bot­tom line is, on good gov­ern­ment grounds, we have to have both vot­er ID laws and absen­tee bal­lot laws.”

Fund and for­mer Bush-era Jus­tice Depart­ment offi­cial and cur­rent Her­itage Foun­da­tion fel­low Hans von Spakovsky have a book com­ing out next week titled Who’s Count­ing: How Fraud­sters and Bureau­crats Put Your Vote at Risk. Fund offered a pret­ty dire pre­dic­tion of how things will go at the polls in a few months.

“The thing I fear most is that we’ll have five, six or sev­en Flori­das this Novem­ber, where the elec­tion is close enough that there’s no longer a mar­gin of vic­to­ry you’re talk­ing about, there’s a mar­gin of lit­i­ga­tion, and we will not know the win­ner of the pres­i­den­tial race for hours or days or per­haps in the case of some­where like Flori­da, per­haps for quite some­time, and the next pres­i­dent takes office or con­tin­ues in office under a cloud,” Fund said.

Fund and von Spakovsky attacked a report from the Bren­nan Cen­ter which claims that mil­lions of vot­ers lack valid pho­to iden­ti­fi­ca­tion and will be pre­vent­ed from vot­ing in Novem­ber.

...

Five or six Floridas...great, that’s all we need.

The new book com­ing out about vot­er fraud by Fund and for­mer Jus­tice Depart­ment Civ­il Rights attor­ney Hans von Spakovsky looks like the kind of “analy­sis” that one might expect from these two. But it’s also quite note­wor­thy that the pair seem to be push­ing the meme that absen­tee vot­er fraud is the pri­ma­ry means of mas­sive vot­er fraud avail­able today. Now what about that issue of a third of US vot­ers now using eas­i­ly-hack­able elec­tron­ic vot­ing machines? Wha’ts their take on that touchy top­ic? Fund actu­al­ly ded­i­cates an entire chap­ter to elec­tron­ic vot­ing machines...in his 2004 book “Steal­ing Elec­tions: How Vot­er Fraud Threat­ens Our Democ­ra­cy”. Just use the “Search inside his book” func­tion and do a search for “Diebold”. Grant­ed, Fund down­plays the threat of elec­tron­ic vot­ing machines, but at least the top­ic is men­tioned in a book about vot­er fraud. Jump ahead to Fund’s new book Who’s Count­ing?: How Fraud­sters and Bureau­crats Put Your Vote at Risk and there does­n’t appear to be a sin­gle ref­er­ence to elec­tron­ic vot­ing machines in the entire thing. Once again, use the “Search Inside this book” option and so some search­es for “Diebold” or even “elec­tron­ic”. Noth­ing (at least noth­ing I could find).

This is all just a reminder that the GOP does­n’t sim­ply ben­e­fit from a sup­pressed minority/poor/elderly vote as a result of all these new vot­er ID laws get­ting put in place across the US. The new laws and the pub­lic fight over them also help dif­fuse any pub­lic atten­tion that might have been paid towards the still grow­ing threat of unse­cure elec­tron­ic vot­ing sys­tems. So if the “five or six Flori­das” sce­nario emerges in Novem­ber, the entire elec­torate and media com­plex is now primed to imme­di­ate­ly jump to “Absen­tee vot­er fraud!” or “Ille­gal immi­grants vot­ing!” instead of the obvi­ous cul­prit of elec­tron­ic vot­ing fraud.

This, of course, is not to say that good ol’ fash­ioned vot­er-sup­pre­sion won’t do the trick. Not at all:

August 6, 2012 7:14 AM
Despite polls, Rom­ney camp sees oppor­tu­ni­ty in Pa.

(CBS News) Though both sides are adept at putting the most favor­able spin on their elec­toral prospects, strate­gists for the Oba­ma and Rom­ney cam­paigns large­ly agree on which so-called swing states are tru­ly up for grabs and which ones clear­ly lean a par­tic­u­lar way.

But per­haps the most glar­ing excep­tion to that broad con­sen­sus is Penn­syl­va­nia, where the two camps are work­ing under wide­ly diver­gent assess­ments of how the race is shap­ing up.

Despite earnest assur­ances that they are tak­ing noth­ing for grant­ed, a com­mon­ly held view among the pres­i­den­t’s re-elec­tion team is that the Key­stone State is all but in the bag. Rom­ney’s team, on the oth­er hand, has long been eye­ing it as a real­is­tic and poten­tial­ly deci­sive pick­up for the Repub­li­can chal­lenger.

“When you talk about Penn­syl­va­nia, the Oba­ma cam­paign is going to roll their eyes,” said Rom­ney’s polit­i­cal direc­tor, Rich Bee­son. “They don’t know it, but it’s rot­ten under­neath for them.”

At first glance, there is lit­tle empir­i­cal evi­dence to back up that claim.

The pres­i­dent has been ahead in every Penn­syl­va­nia poll since Rom­ney became the GOP nom­i­nee, and he leads the Repub­li­can chal­lenger by sev­en per­cent­age points in the lat­est RCP polling aver­age.

...

But what many Repub­li­cans expect to be their strongest card in Penn­syl­va­nia is one that they are often reluc­tant to tout: a new vot­er iden­ti­fi­ca­tion law that requires a spe­cif­ic form of iden­ti­fi­ca­tion in order to cast a bal­lot.

The law is con­sid­ered the most strin­gent of its kind in any swing state in the nation and could cut sig­nif­i­cant­ly into Oba­ma’s mar­gins in the state’s largest city.

“The bot­tom line is Philadel­phia is incred­i­bly impact­ed by that law, and it could have a real effect,” said KDKA-Pitts­burgh polit­i­cal edi­tor Jon Delano, a long­time ana­lyst of Penn­syl­va­nia pol­i­tics.

Delano, who con­sid­ers the state to be “very much in play,” said that there are no reli­able pro­jec­tions on the extent to which the vot­er ID law might depress turnout among Oba­ma vot­ers, but there is no doubt that Repub­li­cans see it as a major boon to Rom­ney’s hopes.

At a June GOP ral­ly, House Major­i­ty Leader Mike Turzai sparked out­rage among vot­ing-rights pro­po­nents when he trum­pet­ed the new law as an accom­plish­ment that would pay polit­i­cal div­i­dends to the Repub­li­can nom­i­nee.

“Vot­er ID, which is going to allow Gov. Rom­ney to win the state of Penn­syl­va­nia: done,” Turzai said, elic­it­ing cheers from the par­ti­san crowd.

...

Ack, the link to their new book in the above com­ment is bro­ken. Here’s a fixed one. Seri­ous­ly, can any­one find a ref­er­ence to ele­cron­ic vot­ing machines in that entire book using the “Search Inside this book” Ama­zon func­tion? Is there a search term that I’m just not think­ing of that will bring up at least one ref­er­ence to elec­tron­ic vot­ing machines? John Fund and Hans von Spakovsky are foot sol­diers in the far-right agen­da so it’s no sur­prise if they want to down­play the risk of elec­tron­ic vot­ing machine rig­ging, but it’s quite a warn­ing sign if a brand new book about vot­ing theft does­n’t even men­tion elec­tron­ic vot­ing machines at least once.

Vir­gini­a’s attor­ney gen­er­al Ken Cuc­cinel­li is now assert­ing that an inves­ti­ga­tion is need­ed to deter­mine whether or not Oba­ma stole the elec­tion through vot­er fraud. He specif­i­cal­ly cites the fact that the Pres­i­dent lost every state with one of the pho­to-ID laws that the GOP had been pin­ing for in states across the coun­try over the last year. And you know what, he’s right. An inves­ti­ga­tion into the need for these new vot­ing vot­er ID laws and the oth­er new GOP-inspired vot­ing laws from this past year would be a won­der­ful idea.

Here’s a ques­tion that’s prob­a­bly worth ask­ing. Repeat­ed­ly and loud­ly:
Why is the Trump cam­paign so ter­ri­fied of the Michi­gan recount?:

Alter­Net

Why Is Trump Fight­ing So Hard to Stop the Recount in Michi­gan?
A mere 0.22 per­cent sep­a­rates the can­di­dates, and thou­sands of bal­lot machines record­ed no pres­i­den­tial pick.

By Steven Rosen­feld / Alter­Net
Decem­ber 1, 2016

The Trump cam­paign filed a legal peti­tion Thurs­day to stop Michigan’s pres­i­den­tial recount, say­ing Jill Stein has no chance of win­ning and no griev­ance, and there’s no way a hand count can be done before the Elec­toral Col­lege meets Decem­ber 19.

“Despite being a blip on the elec­toral radar, [Green Par­ty nom­i­nee] Stein has now com­man­deered Michigan’s elec­toral process,” said Trump’s peti­tion to Michigan’s Board of State Can­vass­er. “Indeed, on the basis of noth­ing more than spec­u­la­tion, Stein asks that Michi­gan res­i­dents endure an expen­sive, time-con­sum­ing recount, and the scruti­ny and hard­ship that comes with it.”

The Trump cam­paign called the recount an “elec­toral farce,” ignor­ing that it has been filed in a state where its can­di­date leads Hillary Clin­ton by almost 11,000 out of 4.8 mil­lion votes cast, the small­est mar­gin of the final three states that gave Trump an Elec­toral Col­lege major­i­ty after elec­tion night. (The Green Par­ty is also seek­ing recounts in Wis­con­sin and Penn­syl­va­nia. Wis­con­sin began its recount Thurs­day, while chal­lenges con­tin­ue in Penn­syl­va­nia as the Green Par­ty is fil­ing local peti­tions at coun­ty elec­tion boards.)

“She does not allege, let alone explain, how a fourth-place fin­ish­er could be ‘aggriev­ed’ by the elec­tion can­vass,” Trump’s peti­tion said, say­ing can­di­dates must be harmed to seek a recount under Michi­gan law. “And even if that could be over­looked, Stein’s request would have to be denied because no recount can be reli­ably com­pet­ed in the time required by state and fed­er­al law” (to meet the Elec­toral College’s dead­line).

Trump’s objec­tion comes just one day after state offi­cials said they would do a full, statewide hand recount of all votes cast in Michi­gan, begin­ning Fri­day, Decem­ber 2. Because of the fil­ing, the state can­vassers’ board, an appoint­ed body that has already pledged its sup­port of the recount, will have to post­pone count­ing until it holds a hear­ing and issues a rul­ing with­in five days. If the recount goes for­ward, Trump’s objec­tions will have delayed the process up to a week and made it that much hard­er to com­plete before pres­i­den­tial elec­tors con­vene.

“You won­der why they are fight­ing it so hard,” said Bob Fitrakis, an Ohio-based attor­ney who was involved in that state’s 2004 pres­i­den­tial recount and has been advis­ing the Green Par­ty in 2016. “Michi­gan, with its thou­sands of under­votes [no vote record­ed for pres­i­dent] in Demo­c­ra­t­ic strongholds—these peo­ple vot­ed the whole tick­et but left off Hillary Clin­ton?”

“They are try­ing to run out the clock,” he said, say­ing that same delay­ing tac­tic was used in Ohio in 2004 when the Green and Lib­er­tar­i­an par­ties filed for a pres­i­den­tial recount in the state where George W. Bush unex­pect­ed­ly beat John Ker­ry, despite media exit polls and oth­er indices sug­gest­ing the incum­bent pres­i­dent would be defeat­ed.

...

On Mon­day, the Michi­gan Board of State Can­vassers cer­ti­fied that Trump won by 10,704 votes, or a mar­gin of just 0.22 per­cent of the total vote. But com­put­er sci­en­tists and elec­tion experts have raised seri­ous con­cerns about elec­tion results in papers filed for the Wis­con­sin recount. These include the vul­ner­a­bil­i­ty of vot­ing machines that can be breached with­out detec­tion and have a ten­den­cy to mis­read bal­lot mark­ings. In Michi­gan, there were 75,335 under-count tallies—votes that machines did not record as select­ing any­one for president—nearly dou­ble the amount record­ed in 2012.

“Amer­i­ca’s vot­ing machines and opti­cal scan­ners are prone to errors and sus­cep­ti­ble to out­side manip­u­la­tion,” said J. Alex Hal­der­man, one of the nation’s lead­ing cyber secu­ri­ty experts and a pro­fes­sor of com­put­er sci­ence at the Uni­ver­si­ty of Michi­gan who filed court papers sup­port­ing a recount.

“Paper bal­lots, like those used in Michigan’s elec­tions, are the best defense we have against cyber­at­tacks, but that defense is only effec­tive if we actu­al­ly look at the paper trail after the elec­tion,” he said. “That’s pre­cise­ly why we need this recount—to exam­ine the phys­i­cal evi­dence, to look under the hood. A recount is the best way, and indeed the only way in 2016, to ensure pub­lic con­fi­dence that the results are accu­rate, authen­tic, and untaint­ed by out­side inter­fer­ence.”

“On Mon­day, the Michi­gan Board of State Can­vassers cer­ti­fied that Trump won by 10,704 votes, or a mar­gin of just 0.22 per­cent of the total vote. But com­put­er sci­en­tists and elec­tion experts have raised seri­ous con­cerns about elec­tion results in papers filed for the Wis­con­sin recount. These include the vul­ner­a­bil­i­ty of vot­ing machines that can be breached with­out detec­tion and have a ten­den­cy to mis­read bal­lot mark­ings. In Michi­gan, there were 75,335 under-count tallies—votes that machines did not record as select­ing any­one for president—nearly dou­ble the amount record­ed in 2012.”

Oh look at that, it turns out there were over 75,000 “under-count” tal­lies, where there’s no vote for a pres­i­den­tial can­di­date (or it did­n’t get record­ed for some rea­son), in a race Trump won by less than 11,000 votes. Might that have some­thing to do with the Trump team’s recount trep­i­da­tion? It seems very pos­si­ble, although if you lis­ten to the Michi­gan state GOP, which filed its own law­suit to stop the recount on Mon­day, the con­cern is over wast­ed tax pay­er mon­ey. Also, the recount will give vot­ers less con­fi­dence that their vote was accu­rate­ly count­ed. Yes, accord­ing to the state GOP, it’s appar­ent­ly a waste of mon­ey when a can­di­date wins by 0.22 per­cent of the vote total and any recount will reduce the vot­ers’ faith in the integri­ty of the vote:

The Detroit News

GOP fights Mich. recount with fed­er­al appeal

Jonathan Oost­ing , Detroit News Lans­ing Bureau
7:15 p.m. EST Decem­ber 5, 2016

Lans­ing — The Michi­gan Repub­li­can Par­ty on Mon­day sought to stop the state’s pres­i­den­tial recount just hours after it began, appeal­ing a fed­er­al court order grant­i­ng a request by Green Par­ty can­di­date Jill Stein.

“We do not believe the fed­er­al courts have juris­dic­tion over state’s being able to run their own elec­tion process,” par­ty Chair­woman Ron­na Rom­ney McDaniel told The Detroit News.

Par­ty attor­neys filed a notice of appeal to the 6th U.S. Cir­cuit Court of Appeals on Mon­day after­noon and were prepar­ing to make the case for “absten­tion,” argu­ing U.S. Dis­trict Judge Mark Gold­smith should not have inter­vened by order­ing the recount to start with legal chal­lenges pend­ing before the Michi­gan Court of Appeals.

Those chal­lenges would have pre­vent­ed offi­cials from begin­ning the recount until at least Wednes­day.

“There’s absolute­ly no rea­son Jill Stein should be grant­ed a recount in our state that’s going to cost tax­pay­ers mil­lions of dol­lars, and the courts should have been able to make that deter­mi­na­tion before the fed­er­al courts inter­vened,” Rom­ney McDaniel said.

Recount work­ers in Ing­ham and Oak­land coun­ties began hand count­ing bal­lots around noon Mon­day, less than 12 hours after Gold­smith ruled that delay­ing the process could stop the state from final­iz­ing its vote total by a Dec. 13 dead­line, jeop­ar­diz­ing Michigan’s role in the pres­i­den­tial elec­tion.

Six addi­tion­al coun­ties, includ­ing Wayne and Macomb, were expect­ed to begin recounts Tues­day with oth­ers fol­low­ing suit lat­er this week and ear­ly next, accord­ing to a sched­ule released by the Sec­re­tary of State’s Office.

The Michi­gan Repub­li­can Par­ty “real­ly has no basis to file an appeal,” said Stein attor­ney Mark Brew­er, for­mer chair­man of the Michi­gan Demo­c­ra­t­ic Par­ty. “The judge wrote a very strong opin­ion, but we’ll respond when they make their argu­ments.”

Repub­li­can Pres­i­dent-elect Don­ald Trump won Michigan’s Nov. 8 elec­tion, top­ping Demo­c­rat Hillary Clin­ton by 10,704 votes, accord­ing to results cer­ti­fied last week by the Board of State Can­vassers. Stein fin­ished a dis­tant fourth, receiv­ing rough­ly 1 per­cent of the vote.

Stein request­ed recounts in Michi­gan and two oth­er states as part of what her cam­paign is call­ing “an elec­tion integri­ty move­ment,” but she has acknowl­edged the effort is unlike­ly to change the out­come of the pres­i­den­tial elec­tion.

Stein cel­e­brat­ed Goldsmith’s order Mon­day in a press con­fer­ence out­side Trump Tow­er in New York, say­ing the judge affirmed the impor­tance of “elec­tion fair­ness and accu­ra­cy, and that these things are the bedrock of our democ­ra­cy.”

Stein has jus­ti­fied her recount request by ques­tion­ing the unusu­al­ly large num­ber of “under­votes” in Michi­gan, where cer­ti­fied results show 75,000 vot­ers who cast a bal­lot on Nov. 8 did not pick a pres­i­den­tial can­di­date.

That “rais­es the pos­si­bil­i­ty that com­mu­ni­ties of col­or have been dis­en­fran­chised by an erro­neous count­ing of the vote,” she said.

Trump sued Michi­gan on Fri­day in an attempt to stop the recount, and the state Court of Appeals is expect­ed to hear argu­ments from his attor­neys and state Attor­ney Gen­er­al Bill Schuette dur­ing a Tues­day after­noon hear­ing in Lans­ing.

“Michi­gan law needs to gov­ern, not some fed­er­al takeover,” Schuette said Mon­day on Fox News.

State courts can­not super­sede Goldsmith’s order that prompt­ed the recount to start, accord­ing to Wayne State Uni­ver­si­ty law pro­fes­sor and legal schol­ar Robert Sedler, but there is a pos­si­bil­i­ty state judges could inval­i­date the recount if they deter­mine Stein was not an “aggriev­ed par­ty” with grounds for the recount request.

But the absten­tion claims at the heart of the Michi­gan Repub­li­can Party’s fed­er­al appeal are “not very strong” because count­ing votes is a fun­da­men­tal right that Gold­smith said could be jeop­ar­dized.

“The worst that can hap­pen to the state is that it spends the mon­ey and the recount doesn’t count any­way,” Sedler said. “Under the prece­dents, as I under­stand them, this would not be an appro­pri­ate case for the court to abstain.”

Stein’s attor­neys last week paid $973,250 to cov­er her legal­ly required fees to seek a recount of Michigan’s 6,300 precincts. Sec­re­tary of State Ruth John­son has said the recount cost could a total of $5 mil­lion, leav­ing the state and coun­ty gov­ern­ments on the hook for the remain­ing $4 mil­lion.

Rom­ney McDaniel said courts should reject the recount request to “set a prece­dent in this state that we are no longer going to let our tax­pay­ers be abused.” The GOP appeal is also expect­ed to argue that Stein wait­ed to file her recount until the last minute and there­fore could not seek court inter­ven­tion.

The Nov. 8 elec­tion “was admin­is­tered fair­ly and legal­ly,” Rom­ney McDaniel said. “Every vot­er in Michi­gan should have con­fi­dence that their vote was count­ed. It was cer­ti­fied, and Jill Stein is try­ing to intro­duce chaos into our elec­tion process.”

The state House Elec­tions Com­mit­tee on Tues­day is set to take up new­ly intro­duced leg­is­la­tion that seeks to retroac­tive­ly require Stein to cov­er all recount costs. The bill would require any can­di­date that received less than 5 per­cent of the vote in any elec­tion to pay for full esti­mat­ed costs.

...

“Stein has jus­ti­fied her recount request by ques­tion­ing the unusu­al­ly large num­ber of “under­votes” in Michi­gan, where cer­ti­fied results show 75,000 vot­ers who cast a bal­lot on Nov. 8 did not pick a pres­i­den­tial can­di­date.”

So the Michi­gan GOP just can’t see how an an unusu­al­ly large num­ber of “under­votes” that dwarfs the actu­al spread between the Trump and Clin­ton is a valid rea­son for a recount. How sur­pris­ing. Although con­sid­er­ing all the clam­or­ing about “rig­ging” by the Trump team before the elec­tion it would be sur­pris­ing if one was fool­ish enough to take Trump at his word.

Regard­less, at this point it’s not even clear a mean­ing­ful recount is going to take place despite the recount being under­way. Why? Because it turns out Michi­gan has a curi­ous law about recounts: if the num­ber of bal­lots in the precinct poll books do not exact­ly match those of the vot­ing machine print­out, the entire precinct is removed from the recount and the orig­i­nal counts are used instead. Even if they’re off by a sin­gle bal­lot.

And, lo and behold, in the Demo­c­ra­t­ic strong­hold of Detroit, where they report­ed that 87 opti­cal scan­ners broke on elec­tion day, 59 per­cent of the precincts could be thrown out due to votes being off by a hand full of votes and in many cas­es by one vote. In oth­er words, the poor­er the precinct and the more error prone its vot­ing tech­nol­o­gy is, the less like­ly that that precinct will actu­al­ly be involved in a recount. So, despite the fact that a wave of bro­ken or mis­cal­i­brat­ed opti­cal scan­ners would be exact­ly the kind of sce­nario that one might expect to pro­duce a large num­ber of “under­votes”, it’s look­ing like those Demo­c­ra­t­ic strong­hold precincts might not get recount­ed. At all:

The Detroit News

Half of Detroit votes may be inel­i­gi­ble for recount

Chad Liven­good and Joel Kurth , The Detroit News
2:54 p.m. EST Decem­ber 6, 2016

One-third of precincts in Wayne Coun­ty could be dis­qual­i­fied from an unprece­dent­ed statewide recount of pres­i­den­tial elec­tion results because of prob­lems with bal­lots.

Michigan’s largest coun­ty vot­ed over­whelm­ing­ly for Demo­c­ra­t­ic can­di­date Hillary Clin­ton, but offi­cials couldn’t rec­on­cile vote totals for 610 of 1,680 precincts dur­ing a coun­ty­wide can­vass of vote results late last month.

Most of those are in heav­i­ly Demo­c­ra­t­ic Detroit, where the num­ber of bal­lots in precinct poll books did not match those of vot­ing machine print­out reports in 59 per­cent of precincts, 392 of 662.

Accord­ing to state law, precincts whose poll books don’t match with bal­lots can’t be recount­ed. If that hap­pens, orig­i­nal elec­tion results stand.

“It’s not good,” con­ced­ed Daniel Bax­ter, elec­tions direc­tor for the city of Detroit.

He blamed the dis­crep­an­cies on the city’s decade-old vot­ing machines, say­ing 87 opti­cal scan­ners broke on Elec­tion Day. Many jammed when vot­ers fed bal­lots into scan­ners, which can result in erro­neous vote counts if bal­lots are insert­ed mul­ti­ple times. Poll work­ers are sup­posed to adjust coun­ters to reflect a sin­gle vote but in many cas­es failed to do so, caus­ing the dis­crep­an­cies, Bax­ter said.

Even so, Bax­ter said it’s unlike­ly all 392 of the city’s precincts with mis­matched num­bers will be dis­qual­i­fied from a recount. The city is in con­tact with elec­tions offi­cials at the state of Michi­gan and Bax­ter pre­dict­ed the num­bers will match when the bal­lot box­es are re-opened for the recount, which starts Tues­day in Wayne Coun­ty at Cobo Cen­ter.

“It’s a chal­lenge, but we’re con­fi­dent the bal­lots will match,” Bax­ter said. “I don’t think it’s going to be 100 peHad this pro­gram­ming error been a sim­ple flip — telling the machine to count ‘yes’ votes from ‘no’ bub­bles and vice-ver­sa — and not the bla­tant­ly obvi­ous error it was, the munic­i­pal can­vass would almost cer­tain­ly have cer­ti­fied the results with­out exam­in­ing even one actu­al bal­lot; the Stoughton news­pa­per would be com­ing up with per­fect­ly believ­able rea­sons why theirs was the first city ever to vote 75% no instead of 75% yes, and a few ref­er­en­dum back­ers and elec­tion-integri­ty activists would be say­ing “That’s got to be a mis­count” to any­one who would lis­ten –which would not be many.rcent, but it nev­er is with a recount.”

Coun­ty reports obtained by The Detroit News, though, indi­cate can­vassers were pro­vid­ed no expla­na­tion for why the num­bers didn’t add up in those precincts. They cer­ti­fied the results of the elec­tion any­way.

In Detroit, 158 of the 392 precincts with bal­lot dis­crep­an­cies had just one extra bal­lot account­ed for either in the poll book or in the bal­lot box, accord­ing to the Wayne County’s can­vass­ing report.

For sub­ur­ban Wayne Coun­ty, 72 per­cent of the 218 precincts box­es with dis­crep­an­cies in the num­ber of bal­lots were off by one bal­lot.

The oth­er bal­lot dis­crep­an­cies in Detroit and Wayne Coun­ty precincts ranged between two and five bal­lots, accord­ing to the report.

U.S. Dis­trict Judge Mark Gold­smith ordered the hand recount of 4.8 mil­lion bal­lots just after mid­night Mon­day, grant­i­ng a motion in favor of Green Par­ty pres­i­den­tial can­di­date Jill Stein. It began Mon­day in Oak­land and Ing­ham coun­ties and con­tin­ues at 9 a.m. Tues­day in Kala­ma­zoo, Kent, Macomb, Ottawa, Washt­e­naw and Wayne coun­ties.

The recount will begin through­out the rest of Michigan’s 83 coun­ties in phas­es this week.

Repub­li­can Pres­i­dent-elect Don­ald Trump won Michi­gan by a razor-thin mar­gin, 10,704 votes. Pre­sum­ably, Clinton’s best oppor­tu­ni­ty to elim­i­nate that mar­gin rest­ed in find­ing uncount­ed bal­lots in Wayne Coun­ty, which she car­ried by a 2–1 ratio.

Dis­qual­i­fy­ing huge num­bers of precincts would make it “almost impos­si­ble” for the for­mer New York sen­a­tor to make up the votes, said Ernest John­son, a Demo­c­ra­t­ic polit­i­cal activist who worked to get out the vote for Clin­ton.

“It’s a real long-shot now because, if I were look­ing for 10,000 votes, the first place I’d look is Wayne Coun­ty,” John­son said. “That’s a huge prob­lem. ... But if any­thing good comes of this it brings up this prob­lem (with vot­ing machines) that needs to be cor­rect­ed.”

$100K bill for Gene­see Co.

Besides Wayne, Clin­ton car­ried Oak­land, Washt­e­naw, Gene­see, Ing­ham, Kala­ma­zoo, Mar­quette and Muskegon coun­ties.

None had near­ly as many prob­lems as Wayne. But at least 13 of 222 precincts in Gene­see Coun­ty are not bal­anced. More than half of those were in heav­i­ly Demo­c­ra­t­ic Flint, accord­ing to coun­ty can­vass­ing reports. The elec­tion was still cer­ti­fied by its board of can­vassers.

“The trou­ble is there’s too much lenien­cy with the board of can­vassers,” said John Glea­son, Gene­see County’s clerk. “They’re not as strin­gent as need to be because they think it won’t affect the out­come of the elec­tion.”

Glea­son, who esti­mat­ed the recount will cost the cash-strapped coun­ty $100,000, said he expects the num­bers to rec­on­cile when work­ers begin the coun­ty recount on Wednes­day.

“It’s impos­si­ble to tell at this point how many will and how many won’t be re-count­able,” added Joseph Rozell, elec­tions direc­tor for Oak­land Coun­ty.

In the first six hours of Ing­ham County’s recount Mon­day, six of 30 precincts from Lans­ing could not be recount­ed. One of the bal­lot con­tain­ers had a hole in it, mak­ing it sus­cep­ti­ble to tam­per­ing and not recount­able, coun­ty Clerk Barb Byrum said.

Elec­tions offi­cials in Washt­e­naw and Mar­quette coun­ties told The News on Mon­day that votes in all precincts were rec­on­ciled.

Krista Haroutun­ian, chair of the Wayne Coun­ty Board of Can­vassers, said recount work­ers will have to sort out the issue when they encounter these bal­lot box­es where the num­bers don’t rec­on­cile with the elec­tion records.

“When a recount is start­ed, these num­bers could change based on infor­ma­tion dis­cov­ered dur­ing the course of the recount, accord­ing to our pro­ce­dures,” Haroutun­ian said. “You don’t know until you get in the mid­dle of it.”

Two tries to match num­bers

State law spells out a pre­scribed cri­te­ria for deter­min­ing whether a precinct can be recount­ed. Work­ers first check to make sure the num­ber of bal­lots on the seal of the con­tain­er match­es the bal­lots record­ed by work­ers on Elec­tion Day.

“If the seal num­ber match­es, then we know it’s not been tam­pered with,” Rozell said.

If num­bers don’t match, then work­ers can count all of the bal­lots in the precinct twice to see if there was an error made by work­ers on Elec­tion Night.

“If it does not match after the sec­ond count, then it’s not recount­able,” Rozell said.

Entire precincts can be set aside in a recount if the bal­lots are stored in an unap­proved con­tain­er or if it isn’t prop­er­ly sealed.

...

“He blamed the dis­crep­an­cies on the city’s decade-old vot­ing machines, say­ing 87 opti­cal scan­ners broke on Elec­tion Day. Many jammed when vot­ers fed bal­lots into scan­ners, which can result in erro­neous vote counts if bal­lots are insert­ed mul­ti­ple times. Poll work­ers are sup­posed to adjust coun­ters to reflect a sin­gle vote but in many cas­es failed to do so, caus­ing the dis­crep­an­cies, Bax­ter said.”

Well isn’t that love­ly: the way Michi­gan’s recount is set up, if a precinct reports opti­cal scan­ner trou­ble or the machines are prone to jam­ming — two issues that one would think would be a great rea­son for a recount — those precincts are left out of the recount. And the prob­lem pre­sum­ably nev­er gets fixed.

So, with all that in mind, let’s take a walk down opti­cal scan­ner under-count­ing mal­func­tion mem­o­ry-lane. opti­cal scan­ner under­count­ing mal­func­tion recent-mem­o­ry-lane:

BradBlog.com

WI Paper Bal­lot Scan­ners Failed to Count 1000s of Votes in ‘Cit­i­zens Unit­ed’ Bal­lot Ref­er­en­dum

Pop­u­lar, oft-mal­func­tion­ing com­put­er tab­u­la­tor used in WI, many oth­er states, tal­lied just 16 votes out of 5,350 cast in Stoughton, WI...
UPDATED: Hand-count finds 4,440 YES votes, 992 NO votes...

By Brad Fried­man on 11/9/2014, 9:05am PT

Though some 5,350 vot­ers are known to have vot­ed in the city of Stoughton in Dane Coun­ty, Wis­con­sin on Tues­day, just 16 of those vot­ers were inter­est­ed in vot­ing in a local bal­lot ref­er­en­dum call­ing for an amend­ment to the U.S. Con­sti­tu­tion to help over­turn the infa­mous Cit­i­zens Unit­ed deci­sion deci­sion — at least accord­ing to the results report­ed by paper bal­lot opti­cal-scan com­put­er tab­u­la­tors there.

“A mal­func­tion with the vot­ing machines in Stoughton Tues­day led to an incom­plete out­come of the city’s ref­er­en­dum on whether to amend the U.S. Con­sti­tu­tion, Stoughton clerk Lana Kropf said,” accord­ing to a terse and some­what cryp­tic report in the Wis­con­sin State Jour­nal on Thurs­day.

The city’s ES&S DS200 paper bal­lot opti­cal-scan­ners (a com­put­er tab­u­la­tion sys­tem plagued with prob­lems in many states over the years) report­ed zero votes for the ini­tia­tive in five of the city’s six vot­ing wards, and just 16 votes (7 Yes, 9 No) in the oth­er.

“Nev­er in my years work­ing in clerks’ offices have I seen some­thing like this,” Kropf told the Jour­nal.

The ini­tia­tive in ques­tion was added to the bal­lot after cit­i­zens gath­ered enough sig­na­tures last July to have it includ­ed on the Novem­ber bal­lot. Accord­ing to the word­ing of the mea­sure, it seeks “to reclaim democ­ra­cy from the expan­sion of cor­po­rate per­son­hood rights and the cor­rupt­ing influ­ence of unreg­u­lat­ed polit­i­cal con­tri­bu­tions and spend­ing.” It is sim­i­lar to ref­er­en­da passed over­whelm­ing­ly in oth­er WI com­mu­ni­ties.

...

She adds: “For­tu­nate­ly, no one believes the city’s vot­ing machines”...

The ref­er­en­dum was based on a Move to Amend cam­paign call­ing for a 28th Amend­ment to the U.S. Con­sti­tu­tion. The ver­sion of the Amend­ment on Stoughton’s bal­lot reads:

1. Only human beings — not cor­po­ra­tions, lim­it­ed lia­bil­i­ty com­pa­nies, unions, non-prof­it orga­ni­za­tions, or sim­i­lar asso­ci­a­tions and cor­po­rate enti­ties — are endowed with con­sti­tu­tion­al rights, and
2. Mon­ey is not speech, and there­fore reg­u­lat­ing polit­i­cal con­tri­bu­tions and spend­ing is not equiv­a­lent to lim­it­ing polit­i­cal speech.

The city of Stoughton’s mea­sure was the only item to appear on the back of the hand-marked paper bal­lots in Stoughton last week. But is it pos­si­ble that only 16 vot­ers out of more than 5,000 turned over the bal­lot to notice it and/or had an opin­ion on the res­o­lu­tion either way?

Well, the good news is, we now appear to know what actu­al­ly hap­pened and should be able to deter­mine what last Tues­day’s vot­ers real­ly thought about the ref­er­en­dum.

The bad news is, what­ev­er caused this prob­lem could eas­i­ly have affect­ed any num­ber of oth­er elec­tions in any num­ber of oth­er places both in Wis­con­sin and else­where where iden­ti­cal and sim­i­lar tab­u­la­tion sys­tems are used. With­out a hand-count of those bal­lots, it’s just as like­ly that nobody would ever notice a sim­i­lar prob­lem.

McKim of the pro­gres­sive Wis­con­sin Grass­roots Net­work offered a ter­rif­ic piece explain­ing the prob­lem and why it appears to have hap­pened in Stoughton, accord­ing to local elec­tion offi­cials:

The Dane Coun­ty clerk’s office sets up ES&S DS200 opscan vot­ing machines for each munic­i­pal­i­ty before every elec­tion. To do that, they pro­gram bal­lot-read­ing instruc­tions onto ‘mem­o­ry sticks,’ which look like (and may be for all this non-IT per­son knows) flash dri­ves. DS200 bal­lots are marked along the edges with black index marks. Each of the bub­bles where votes are marked is close to one of these index marks. Trans­lat­ed to Eng­lish from com­put­er-ese, the instruc­tions say, “If the bub­ble clos­est to the top index mark on the right edge of the front of the bal­lot is filled in, count a vote for Can­di­date Smith. If the bub­ble clos­est to the sec­ond-from-the-top index mark is filled in, count a vote for Can­di­date Jones.

When Stoughton’s bal­lots came back from the print­ers, the ref­er­en­dum was per­fect­ly placed on the bal­lot right where it was intend­ed to be. How­ev­er, the per­son who pro­grammed the sticks wrote instruc­tions that told the machine to look at the areas two index marks below the ref­er­en­dum, in white space. That explains the thou­sands of appar­ent­ly blank bal­lots.

But what explains the 16 votes? Did 16 of Stoughton’s vot­ers know they could get their vote count­ed by mak­ing a black mark an inch below the bub­bles that every­one else was fill­ing in? Nope. Truth is, they were vot­ing so enthu­si­as­ti­cal­ly for some­one on the front of the bal­lot that their votes were bleed­ing through the paper into the white space below the ref­er­en­dum on the back.

Bleed-though usu­al­ly isn’t a prob­lem, Dane Coun­ty Clerk Scott McDonell told me today, because when two-sided bal­lots are designed, the bub­bles are placed so that any bleed-through from par­tic­u­lar­ly heav­i­ly marked votes will fall into white space on the oth­er side–as these marks did. Usu­al­ly, the machines are not look­ing for votes in the white spaces. Not usu­al­ly, but they did on Tues­day in Stoughton.

Pre­sum­ing the expla­na­tion from offi­cials is accu­rate, McKim goes on to spec­u­late about a few poten­tial rea­sons it might have occurred in Stoughton:

* The pre­vi­ous vet­er­an City Clerk resigned around the time that the peti­tions were hand­ed in and the mea­sure was added to the bal­lot and some­one less expe­ri­enced took over;
* The late addi­tion to the bal­lot and con­fu­sion about WI Repub­li­cans struck-down, then back on, then back off Pho­to ID vot­ing restric­tions may have exac­er­bat­ed the prob­lem by eat­ing in to time nor­mal­ly spent proof-read­ing bal­lots and test­ing com­put­er tab­u­la­tors;
* The machines may not have gone through their man­dat­ed pre-elec­tion tests at all, even though the City Clerk claims they did.

That last seems the most like­ly and trou­bling, since it’s dif­fi­cult to see how the issue as explained would not have been dis­cov­ered dur­ing pre-elec­tion Log­ic and Accu­ra­cy (L&A) test­ing. Poor L&A test­ing, or com­plete lack there­of, has been a fair­ly reg­u­lar prob­lem with elec­tron­ic vot­ing and tab­u­la­tion sys­tems over the years in the U.S.

As McKim notes, while the City Clerk is claim­ing test­ing was done, “No cit­i­zens were present at the pub­lic pre-elec­tion vot­ing machine test to ensure that it was done prop­er­ly.” She fur­ther explains: “I don’t see any way that test could have been con­duct­ed prop­er­ly and not dis­cov­ered the faulty pro­gram­ming.”

While the Wis­con­sin Grass­roots Net­work did observe pre-elec­tion test­ing in oth­er towns, McKim says that nobody did so in Stoughton. “Cit­i­zens have got to do our bit,” she appro­pri­ate­ly cau­tions. “We can­not relent­less­ly demand trans­paren­cy in gov­ern­ment and relent­less­ly fail to show up for things like pub­lic demon­stra­tions of the vot­ing machines’ abil­i­ty to pro­duce — or in this case, not pro­duce — an error-free count.”

For­tu­nate­ly, the prob­lem this time in Stoughton — whether the result of mal­func­tion, malfea­sance or mis­fea­sance — was quick­ly noticed and should soon be eas­i­ly rec­ti­fied. A hand-count of those hand-marked paper bal­lot will take place on Mon­day to deter­mine the real results of the ref­er­en­dum. Luck­i­ly, the town votes on paper. If it did­n’t, there would be no way to ever deter­mine the actu­al intent of the vot­ers.

But the issue, even as explained above, also reveals once again how easy it is to “trick” these sys­tems — even com­put­er-tab­u­lat­ed paper-bal­lots sys­tems — via sim­ple, stan­dard bal­lot pro­gram­ming procedures...not that any­one would ever do such a thing either acci­den­tal­ly or on pur­pose.

The very same failed ES&S DS200 tab­u­la­tion sys­tems are used in Flori­da, Ohio, New York and oth­er states in addi­tion to Wis­con­sin. As The BRAD BLOG report­ed in 2012 after the com­pa­ny’s man­u­fac­tur­er ES&S con­firmed some of the prob­lems that appear to have led to thou­sands of votes being mistal­lied in New York’s 2010 elec­tion, “Over 4 mil­lion reg­is­tered vot­ers in FL, more than 6.5 mil­lion in NY, over 1 mil­lion in OH, and vot­ers in more than 50 WI municipalities...are set to have their votes tal­lied — accu­rate­ly or oth­er­wise — by the failed ES&S DS200” dur­ing that year’s Pres­i­den­tial elec­tion.

Opti­cal-scan paper bal­lot sys­tems made by oth­er man­u­fac­tur­ers have failed in sim­i­lar and even more spec­tac­u­lar fash­ion. Occa­sion­al­ly it gets noticed, as it did, for exam­ple, in Palm Beach Coun­ty, FL’s munic­i­pal elec­tions in March of 2012, when, luck­i­ly, a sharp-eyed elec­tion offi­cial noticed that the op-scan paper bal­lot sys­tem declared incor­rect “win­ners” in four dif­fer­ent races. Only a hand-count there was able to deter­mine the real win­ners of the con­tests. (The man­u­fac­tur­er of that sys­tem, used in dozens of states, was forced to admit the fail­ure was due to a “bug” present in all of the com­pa­ny’s vot­ing and tab­u­la­tion sys­tems — many of which are also used in Wis­con­sin.)

“If the hand count goes well on Mon­day,” McKim notes (pre­sum­ing the chain of cus­tody of the hand-marked paper bal­lots in Stoughton has been secure since Elec­tion Day), “the munic­i­pal ref­er­en­dum results will be final and cer­ti­fied — and more guar­an­teed-accu­rate than any­thing else on the bal­lot.”

She then offers this impor­tant, cau­tion­ary, spot-on reminder:

Had this pro­gram­ming error been a sim­ple flip — telling the machine to count ‘yes’ votes from ‘no’ bub­bles and vice-ver­sa — and not the bla­tant­ly obvi­ous error it was, the munic­i­pal can­vass would almost cer­tain­ly have cer­ti­fied the results with­out exam­in­ing even one actu­al bal­lot; the Stoughton news­pa­per would be com­ing up with per­fect­ly believ­able rea­sons why theirs was the first city ever to vote 75% no instead of 75% yes, and a few ref­er­en­dum back­ers and elec­tion-integri­ty activists would be say­ing “That’s got to be a mis­count” to any­one who would lis­ten –which would not be many.

Bin­go.

[Hat-tip WI elec­tion integri­ty champ John Wash­burn]

* * *

UPDATE 11/11/2014: While the Lana Kropf, the City Clerk of Stoughton, has not respond­ed to our email queries, the Dane Coun­ty Clerk’s elec­tion web­site has final­ly been updat­ed with some new num­bers fol­low­ing the hand-count of the ref­er­en­dum which was sup­posed to begin on Mon­day...

They show the ref­er­en­dum eas­i­ly pass­ing 4440 to 992. It’s unclear why Wards 9 through 11 show zero votes, how­ev­er. If we can learn more, we’ll update again.

“But the issue, even as explained above, also reveals once again how easy it is to “trick” these sys­tems — even com­put­er-tab­u­lat­ed paper-bal­lots sys­tems — via sim­ple, stan­dard bal­lot pro­gram­ming procedures...not that any­one would ever do such a thing either acci­den­tal­ly or on pur­pose.”

Yep, it turns out paper bal­lots aren’t as fool­proof as they might seem because they still have to be read by an opti­cal scan­ner. And once that scan­ner gets intro­duced into the equa­tion, all the fun “bugs” that come from elec­tron­ic vot­ing machines can poten­tial­ly get intro­duced to the paper bal­lot count too. Which, of course, it why it should be con­sid­ered absolute­ly nec­es­sary to hand count paper bal­lots in a very close elec­tion. Espe­cial­ly in precincts that report­ed opti­cal scan­ner issues on elec­tion day. And, of course, the Trump team dis­agrees.

In oth­er news...

There was a report put out last week about why vot­ing machines fail that makes an impor­tant point about the vot­ing machine indus­try that’s rarely brought up that direct­ly relates to the grow­ing vot­ing machine fias­co in Geor­gia involv­ing the state’s new vot­ing machines: The indus­try is basi­cal­ly an oli­gop­oly dom­i­nat­ed by Elec­tion Sys­tems and Soft­ware (ES&S), Domin­ion Vot­ing Sys­tems, and Hart Inter­Civic. And as a result there’s very lit­tle actu­al inno­va­tion in the indus­try an indus­try known for one ‘glitch’ and ‘hic­cup’ after anoth­er. That’s one of the find­ings in a new report by The Markup — a new inves­tiga­tive jour­nal­ism out­fit start­ed by a pair of ProP­ub­li­ca jour­nal­ists — that explored by vot­ing machines tend to fail on elec­tion day. And it’s a find­ing that the indus­try itself did­n’t entire­ly dis­pute, with a vice pres­i­dent of mar­ket­ing at Hart Inter­Civic acknowl­edg­ing that it’s rare to see new com­pa­nies enter the mar­ket, which he blamed on due to reg­u­la­tion and the fact that gov­ern­ments rarely buy new equip­ment. Those three com­pa­nies large­ly split a mar­ket that’s described as hav­ing a rel­a­tive­ly mod­est $300 mil­lion in annu­al rev­enues.

So there isn’t enough demand to jus­ti­fy a robust com­pet­i­tive mar­ket because the machines are intend­ed to only be updat­ed peri­od­i­cal­ly. If any­thing, all the faulty tech­nol­o­gy has increased sales over the years to replace all the faulty equip­ment, although anoth­er rea­son giv­en in the report for the lack of inno­va­tion in vot­ing machine tech­nol­o­gy is that it can take years and cost mil­lions of dol­lars for new equip­ment to meet state and fed­er­al guide­lines and old­er ven­dors are allowed to cer­ti­fy machines accord­ing to old­er, less rig­or­ous stan­dards. In oth­er words, the indus­try does­n’t think its prof­itable enough to make afford­able secu­ri­ty machines. It’s anoth­er rea­son to rely on tra­di­tion­al meth­ods like paper bal­lots: the vot­ing machine mar­ket­place is a fun­da­men­tal­ly dys­func­tion­al mar­ket­place that can’t pro­duce secure sys­tems:

Mic

Why do vot­ing machines break on Elec­tion Day?

By Adri­anne Jef­fries
Sep. 14, 2020

This arti­cle was orig­i­nal­ly pub­lished by The Markup, a non­prof­it news­room cov­er­ing the ways tech­nol­o­gy is chang­ing soci­ety. Fol­low them on Twit­ter and Face­book, and sub­scribe to their newslet­ter here.

Between wor­ries over a con­test­ed result and warn­ings of chaos come Nov. 3, it’s easy to get an ulcer won­der­ing exact­ly what might hap­pen in the days after the 2020 pres­i­den­tial elec­tion. Much of the cur­rent chat­ter cen­ters on mail-in bal­lots, but mil­lions of peo­ple will vote in per­son this year—and con­fi­dence that their votes will be accu­rate­ly tal­lied by vot­ing machines is crit­i­cal.

Vot­ing machines don’t have the best rep­u­ta­tion, par­tic­u­lar­ly since the “hang­ing chad” issue that left the 2000 pres­i­den­tial elec­tion unde­cid­ed for 36 days. Elec­tron­ic machines were sup­posed to solve those prob­lems, but every elec­tion prompts fresh head­lines about scary tech­nol­o­gy fail­ures. Take the 2020 primaries—a pos­si­ble pre­view of what we might see in Novem­ber: “Cod­ing mishaps” in Iowa;; “bro­ken machines” in Dal­las and New York; “glitch­es” in L.A; in parts of Geor­gia, vot­ers wait­ed as long as eight hours dur­ing the June 9 pri­ma­ry as poll work­ers wres­tled with new vot­ing equip­ment.

Such mishaps have real con­se­quences. Long lines may stop peo­ple from vot­ing and turn them off from par­tic­i­pat­ing in future elec­tions. Stud­ies also show Black and Lati­no vot­ers are more like­ly to face long lines. And vot­ers with dis­abil­i­ties can lose their right to a secret bal­lot if acces­si­ble vot­ing machines don’t work as intend­ed.

But how big a threat to our elec­tions are “bro­ken machines”? That vague and wide­ly used phras­ing actu­al­ly cov­ers a wide range of prob­lems, from major soft­ware and hard­ware break­downs to more mun­dane prob­lems like poll work­ers for­get­ting to plug machines in. But vot­ers should take some com­fort: Most vot­ing machine prob­lems are minor, and with prop­er prepa­ra­tion, none should impair a fair elec­tion.

“Vot­ing Machine” Fail­ures Are Often Human Errors

Elec­tions are com­pli­cat­ed to run, but the most com­mon issues with vot­ing machines tend to be rel­a­tive­ly sim­ple, said Matt Bern­hard, an elec­tion secu­ri­ty researcher who observed elec­tions in Geor­gia in 2018 and Mis­sis­sip­pi in 2019 and worked as a tech­ni­cian dur­ing this March’s pri­ma­ry in Michi­gan.

“You know, it’s sil­ly stuff,” he said. Take one Michi­gan elec­tion, when work­ers at a polling place cooked food through­out the day.

“They had one too many Crock-Pots, and they killed the precinct’s pow­er just because it blew a cir­cuit,” Bern­hard said.

Elec­tron­ic poll­books, typ­i­cal­ly lap­tops or tablets con­nect­ed to vot­er reg­is­tra­tion data­bas­es, are anoth­er fre­quent source of error, he said. In Georgia’s pri­ma­ry, Bern­hard said, a poll work­er mis­tak­en­ly told a vot­er she was at the wrong polling place after putting the poll­book in statewide mode instead of precinct mode.

Touch screen fail­ures, where a vot­er tries to select one can­di­date but some­how picks anoth­er, also pro­duce some of the most dis­con­cert­ing errors for vot­ers; this “vote flip­ping” phe­nom­e­non often hap­pens because poll work­ers neglect­ed to cal­i­brate the machines before vot­ing start­ed, Bern­hard said.

A law­suit filed against state and coun­ty elec­tion offi­cials by Georgia’s Demo­c­ra­t­ic Par­ty in August cit­ed more than three dozen instances of prob­lems with vot­ing equip­ment in elec­tions going back to 2014. Most of the sit­u­a­tions are not described in detail, but the details that do appear are banal: The inter­net went out, the print­ers had no paper, there weren’t enough elec­tri­cal out­lets. The defen­dants filed a motion to dis­miss on Sep­tem­ber 9.

Any of these sit­u­a­tions might be report­ed in the press as “bro­ken machines,” Bern­hard said.

Deep­er Tech­ni­cal Issues Do Hap­pen

In 2010, one precinct in the Bronx threw out an unusu­al­ly high per­cent­age of bal­lots due to vot­ers mark­ing more than one choice. State elec­tion offi­cials inves­ti­gat­ed and found that the machines had added “phan­tom votes” in part because they were over­heat­ing. In 2014, 26 machines in Vir­ginia Beach caused vote flip­ping because the glue in the screen had degrad­ed. In 2006, com­plaints of vote flip­ping in North Lit­tle Rock, Arkansas, were diag­nosed as an opti­cal illu­sion expe­ri­enced only by vot­ers over six feet tall. A report by Los Ange­les Coun­ty, which unveiled a new vot­ing sys­tem dur­ing its 2020 pri­ma­ry, con­clud­ed that 1,297 out of 23,104 vot­ing machines, more than 5 per­cent, had faulty print­ers due to a “man­u­fac­tur­ing defect.”

Vot­ing machine ven­dors and elec­tion offi­cials should be required to report issues with machines, said Lar­ry Nor­den, direc­tor of the Elec­tion Reform Pro­gram at the Bren­nan Cen­ter for Jus­tice.

“There is no real cen­tral repos­i­to­ry for design flaws in sys­tems the way that there would be for auto­mo­biles,” he said, and some issues are nev­er con­clu­sive­ly diag­nosed.

“I still don’t know exact­ly what hap­pened in Geor­gia in the pri­maries,” Nor­den said.

Cyber­se­cu­ri­ty experts have also repeat­ed­ly demon­strat­ed vul­ner­a­bil­i­ties in vot­ing sys­tems that are in use across the coun­try. In 2019, researchers at the DEF CON hack­ing con­fer­ence in Las Vegas com­pro­mised more than 100 vot­ing machines, all of which were cer­ti­fied for use in at least one juris­dic­tion at the time.

The good news is that most vot­ers will vote in a way that gen­er­ates a paper trail, which allows elec­tion offi­cials to check that machines are record­ing votes cor­rect­ly, Nor­den said.

Why Do These Prob­lems Keep Hap­pen­ing?

Experts point­ed to insuf­fi­cient poll work­er train­ing, aging equip­ment, out­dat­ed fed­er­al stan­dards, and a lack of com­pe­ti­tion in the indus­try as the main rea­sons vot­ing machines go down.

The hun­dreds of thou­sands of poll work­ers crit­i­cal to mak­ing vot­ing hap­pen are most­ly tem­po­rary hires. They receive a crash course every year in elec­tion law and equip­ment in order to pull off a com­plex, high-stress, high-stakes event.

Accord­ing to the law­suit in Geor­gia, a precinct cap­tain in Dekalb Coun­ty dur­ing the 2020 pri­ma­ry said he received only “an intro­duc­to­ry course” on how to use the equip­ment that did not include instruc­tion on trou­bleshoot­ing tech­ni­cal issues.

The vot­ing sys­tems them­selves, which are essen­tial­ly made up of com­put­ers, print­ers, and scan­ners, tend to be old.

The 2002 Help Amer­i­ca Vote Act, or HAVA, inspired by the 2000 pres­i­den­tial elec­tion, cre­at­ed nation­al stan­dards for vot­ing equip­ment and incen­tivized states to upgrade.

Many juris­dic­tions still use the now 20-year-old vot­ing machines paid for with the more than $3 bil­lion allot­ted through that act.

Experts also wor­ry about a lack of inno­va­tion in the vot­ing equip­ment mar­ket, which is dom­i­nat­ed by three pri­vate firms—Election Sys­tems and Soft­ware (ES&S), Domin­ion Vot­ing Sys­tems, and Hart InterCivic—splitting a rel­a­tive­ly mod­est $300 mil­lion rev­enue every year, accord­ing to a report spon­sored by the Penn Whar­ton Pub­lic Pol­i­cy Ini­tia­tive.

Cer­ti­fy­ing new equip­ment to meet state and fed­er­al guide­lines can take years and cost mil­lions, accord­ing to the report, deter­ring new com­pa­nies from devel­op­ing bet­ter prod­ucts. Old­er ven­dors, though, are allowed to cer­ti­fy machines accord­ing to old­er, less rig­or­ous stan­dards.

Steven Sock­well, vice pres­i­dent of mar­ket­ing at Hart Inter­Civic, acknowl­edged that it’s rare to see new com­pa­nies enter the space due to reg­u­la­tion and the fact that gov­ern­ments rarely buy new equip­ment, but he dis­put­ed the idea that the com­pa­ny is not inno­v­a­tive. Kati­na Granger, pub­lic rela­tions man­ag­er for ES&S, said the com­pa­ny sees “robust com­pe­ti­tion in our mar­ket­place every day.”

Domin­ion did not respond to a request for com­ment.

What Are Some Pos­si­ble Solu­tions?

More fund­ing and new fed­er­al guide­lines would go a long way toward improv­ing the vot­ing expe­ri­ence, but there are some short­er-term solu­tions that could help.

Longer or more inten­sive poll work­er train­ing could help reduce the num­ber of prob­lems on Elec­tion Day.

The major­i­ty of poll work­ers are over 60, so experts expect some of the usu­al work­ers will stay home in the 2020 gen­er­al elec­tion due to the risk of con­tract­ing COVID-19. Recruit­ing new poll work­ers could reduce some prob­lems on Elec­tion Day, elec­tion offi­cials said.

Some states have found ways to reduce the bot­tle­necks that hap­pen when lots of peo­ple all go to vote at the same time. In Ari­zona vot­ers can vote by mail or vote ear­ly, and many coun­ties have “vote cen­ters” that take vot­ers from any precinct.

“If they’re wait­ing in a long line and something’s down and they don’t want to wait, they have the option to go to anoth­er loca­tion,” said Janine Pet­ty, the state’s deputy elec­tions direc­tor. Ari­zona doesn’t see a lot of issues with bro­ken machines, she said.

Per­haps the best solu­tion is to plan on things going wrong. Precincts should have a back­up paper poll­book, plen­ty of paper bal­lots, and a lock­box for bal­lots that can’t be count­ed on the spot.

“Vot­ing sys­tems aren’t per­fect, and that’s why there’s always a plan B, right?” said Tina Bar­ton, the city clerk and elec­tions admin­is­tra­tor for Rochester Hills, Michi­gan.

She recalled how dur­ing the March pri­ma­ry, the cor­ner of a voter’s bal­lot tore off and jammed the scan­ner. Stan­dard pro­ce­dure kicked in: Vot­ers dropped their bal­lots into a locked com­part­ment in the machine. Bar­ton brought over a spare machine, and two poll work­ers, a Repub­li­can and a Demo­c­rat, over­saw the trans­fer of votes from the jammed machine to the new one. A tech­ni­cian arrived short­ly after and repaired the jammed machine on site so that it could be deployed again if need­ed. No votes were lost, and vot­ers were only delayed by a few min­utes, Bar­ton said—and if they hadn’t been able to replace the machine, the bal­lots from the aux­il­iary bin would have been hand-count­ed at the board of can­vassers.

...

————

“Why do vot­ing machines break on Elec­tion Day?” by Adri­anne Jef­fries; Mic; 09/14/2020

“Vot­ing machines don’t have the best rep­u­ta­tion, par­tic­u­lar­ly since the “hang­ing chad” issue that left the 2000 pres­i­den­tial elec­tion unde­cid­ed for 36 days. Elec­tron­ic machines were sup­posed to solve those prob­lems, but every elec­tion prompts fresh head­lines about scary tech­nol­o­gy fail­ures. Take the 2020 primaries—a pos­si­ble pre­view of what we might see in Novem­ber: “Cod­ing mishaps” in Iowa;; “bro­ken machines” in Dal­las and New York; “glitch­es” in L.A; in parts of Geor­gia, vot­ers wait­ed as long as eight hours dur­ing the June 9 pri­ma­ry as poll work­ers wres­tled with new vot­ing equip­ment.”

Ever since the 2000 Bush v Gore deba­cle that prompt­ed Con­gress to thrust elec­tron­ic vot­ing machines on the states it’s been a US elec­toral rit­u­al of watch­ing the var­i­ous ‘glitch­es’ and ‘train­ing prob­lems’ or ‘iso­lat­ed inci­dents’ with Amer­i­ca’s vot­ing sys­tems dur­ing the pri­maries that may or may not be the same ‘glitch­es’ and ‘train­ing prob­lems’ or ‘iso­lat­ed inci­dents’ that emerge on elec­tion day. And when indus­try fix­es the prob­lems there’s always new prob­lems. Some­times they’re new new prob­lems and some­times it’s the same old new prob­lems. What nev­er changes is an that the indus­try is only mild­ly inter­est­ed in secu­ri­ty at best because it’s not prof­itable enough for them to care more:

...
Why Do These Prob­lems Keep Hap­pen­ing?

Experts point­ed to insuf­fi­cient poll work­er train­ing, aging equip­ment, out­dat­ed fed­er­al stan­dards, and a lack of com­pe­ti­tion in the indus­try as the main rea­sons vot­ing machines go down.

...

The vot­ing sys­tems them­selves, which are essen­tial­ly made up of com­put­ers, print­ers, and scan­ners, tend to be old.

The 2002 Help Amer­i­ca Vote Act, or HAVA, inspired by the 2000 pres­i­den­tial elec­tion, cre­at­ed nation­al stan­dards for vot­ing equip­ment and incen­tivized states to upgrade.

Many juris­dic­tions still use the now 20-year-old vot­ing machines paid for with the more than $3 bil­lion allot­ted through that act.

Experts also wor­ry about a lack of inno­va­tion in the vot­ing equip­ment mar­ket, which is dom­i­nat­ed by three pri­vate firms—Election Sys­tems and Soft­ware (ES&S), Domin­ion Vot­ing Sys­tems, and Hart InterCivic—splitting a rel­a­tive­ly mod­est $300 mil­lion rev­enue every year, accord­ing to a report spon­sored by the Penn Whar­ton Pub­lic Pol­i­cy Ini­tia­tive.

Cer­ti­fy­ing new equip­ment to meet state and fed­er­al guide­lines can take years and cost mil­lions, accord­ing to the report, deter­ring new com­pa­nies from devel­op­ing bet­ter prod­ucts. Old­er ven­dors, though, are allowed to cer­ti­fy machines accord­ing to old­er, less rig­or­ous stan­dards.

Steven Sock­well, vice pres­i­dent of mar­ket­ing at Hart Inter­Civic, acknowl­edged that it’s rare to see new com­pa­nies enter the space due to reg­u­la­tion and the fact that gov­ern­ments rarely buy new equip­ment, but he dis­put­ed the idea that the com­pa­ny is not inno­v­a­tive. Kati­na Granger, pub­lic rela­tions man­ag­er for ES&S, said the com­pa­ny sees “robust com­pe­ti­tion in our mar­ket­place every day.”
...

It’s all a time­ly reminder that the Repub­li­can gov­ern­ing phi­los­o­phy of mass pri­va­ti­za­tion of gov­ern­ment ser­vices includes the require­ment that either the ser­vice is run in a high­ly prof­itable man­ner or the ser­vice sucks, typ­i­cal­ly both. And that applies to ser­vices as fun­da­men­tal as vot­ing infra­struc­ture.

Also keep in mind that if the elec­tron­ic vot­ing machine indus­try faces the hid­den require­ment of installing hid­den back doors for the GOP to manip­u­late the vote that might also pre­vent some new entrants and com­pe­ti­tion.