| Articles Home |
I hear it over and over: "The paper scanner ballot means we are not going to have election fraud by computer manipulation." And how does it mean that? I can think of lots of ways to steal an election with optical scanners and paper ballots. In Minnesota, one of them may have happened in the 2006 Governor race, won by just over a half percent.
In Minnesota ES&S reported lobby expense went from $0-2003, $7000-2004, $40,000-2005. Lobbiest expenses listed at the MN Campaign Finance and Public Disclosure Board. Because of HAVA, in 2005 all Minnesota counties without voting machines adopted ES&S Model 100 precinct-level ballot counters with an Intel processor, and QNX operating system, and PCMCIA memory cards. and ES&S' AutoMARK Voter Assist Terminals for disabled access, except for the few counties that already had Diebold ballot scanners. Diebold reported no lobbying expenses and sold no machines. The contracts got ES&S about $39 million of the $44 million in HAVA (Help America Vote Act) money for Minnesota, not a bad return on a measly $47,000 of registered lobby expense.
Before HAVA most of the counties used hand counted ballots and ran the elections with there own staff. Now most of them do not have the expertise in the staff to run an election, they must contract to use ES&S technical staff. A few of the larger counties, including Diebold customers, do have most of the expertise to run an election yet all still contract with outside technicians.
Since elections are only a couple times a year, the procedures and technical hoo-ha is now too complex to have an election without outside assistance. A central point of election tampering has been inserted into the process. A software "update" by ES&S and most of Minnesota voters can be affected. A single group or even one contracted person now can be involved in many county elections, a basic defense, decentralized elections, is now breached. Ballot scanners run software, all software can be cracked, especially by insiders, a centrally controlled non-public technical staff is in place to install the software and is responsible for "counting".
Yes there is a paper ballot, but how to get to a recount? With these private corporations and equipment between the citizens and the ballots a verified hand count of the paper ballot somehow never happens. Generally in most US state here is what will happen:
-First they usually try to say that they looked at the totals
and they added up in the central counting system.
-Then they look at the individual machine totals and add them up
but no recounts.
-Then they may run the ballots through the machines again, but never
hand count the ballots.
- By now the chain of custody is lost and the mess ends up in court
and the result is there are no ballots counted.
By that time there are many delays and an explosion of actors are in the process, bloggers, newspapers, radio, tv, the various courts (county, state, even fed) ordering various different types of recounts ( but not of actual ballots), declaring different types of audits of software and hardware (worthless noise), the chain of custody of the ballots gets confused or tainted ( an excuse to nullify a real ballot recount) as various paramilitary actors (cops, sheriffs, state police, even feds, (federal marshals, FBI ) and bureaucrats (City, county, state, feds (US Attorney!) ) take or run from responsibility and put different spins on the situation. After a while the issue seems to go away from counting actual ballots and goes to race, religion, guns, abortion, scandal, personal conflicts or whatever.
Usually to complete an action there must be motivation to do it, like say, ensuring a Governor candidate wins in a tight race, the method would be to use a central point of control to manipulate counts, the risk of failure would have to be low. Again, what happens if there are recounts in most states?
-After a million delays and much confusion they may actually count some of the ballots, but that is a very small probability of happening. And if any counting happens it may be limited to a few precincts or counties or certain classes of ballots like "provisional" or "absentee". The person wanting the count has to have a lot of money and political pull to make it happen. Remember, they never did recount paper ballots in Florida in 2000, the US Supreme Court, an actor that should not have been injected into the state election took over and stopped any recount.
So a paper "backup" is almost worthless, it is so rare that a count of actual ballots happens that unless the fraud or error is glaringly huge the recount will not happen. A thumb on the scale changing 1/2%-3% of a vote will sail on through, and that is how you do it, just enough to win, and when you just a slant a few precincts to avoid the 2 to 4 precinct per county random checks you have an excellent chance of getting away with it.
Out of the 3,500 precincts in the state there are not many checked, 4 precincts per congressional district (8) are checked by the statewide canvassing board and each county canvassing board checks at least 2 but no more than 4 precincts. At best 32 (state canvassing board) and 87*2 (minimum per county) + 10*2 (large county additional precincts) or possibly 226 precincts total will be checked of the 3,500, or about 6 1/2%. So this could be a 90+% chance of beating the odds of getting caught in an audit.
According to MN Statute 206.89 POSTELECTION REVIEW OF VOTING SYSTEMS ,(google this), each county is responsible for picking the "random" precincts and doing an audit count. If there are problems over 1/2% 3 more precincts are randomly picked, if again there are more problems over 1/2% the whole county is counted by audit. This is probably good for a county wide programming error or a fraud done in every precinct, but professionals or someone who thinks about this more than 10 minutes and reads the statute may try to avoid an obvious mistake.
For example, to beat the current audit system in Minnesota fix 265 votes in each of 83 precincts, one per each of 83 ES&S counties, about 22,000 votes (what the current Governor won by.) Let's say that the 1/3 of the precincts in the Metro area use Diebold or are the Metro counties that have staff that is too sophisticated to mess with, so we are down to 2333 precincts of which 83*2 (minimum per county) + 3*2 (big counties) + 24 (6 congressional districts), or about 196 precincts are checked. I am back of the envelope guessing the probability of getting caught in at least one county of the 83 is about 8 1/2%, not a sure thing, but not bad either. And the chance of getting caught in 2 or more counties to establish a state wide pattern is very low. So what happens when they catch one county problem, they look at more of the precincts in that county only, no problems found, so they say 'unexplained glitch', the 265 votes in the one precinct caught do not affect the election. The other counties precincts are not found, you still win even when caught.
So if you can influence a few counties canvass boards to "by lot" pick only the clean precincts to audit the odds start to improve over the already impressive 90+%. Jacking 500 or even 1000 or more votes per "cooperative" county starts to look pretty easy. And the damage is contained to only the county that gets caught, if even that. How could the people that set up our elections resist this sure bet. I wonder if they did.
I checked the canvas boards 2006 audit results and yes there are differences in the audited precincts results, usually just a handful of votes, no clear pattern. The notes of how the precincts were chosen by each county are a bit sketchy, however. Most other states do not even have an audit scheme that has a 6-9% chance of finding widespread error or fraud, they just fall apart in a flailing mess of lawsuits, arbitrary actions and confusion.
So, when was the last time you heard of an actual ballot recount? Florida 2000? Ohio 2004? New Mexico 2004? Haw! Hatch lost a close election, how odd he was the only DFL to lose by just a shade over 1/2% (the automatic machine recount limit) in the big DFL win sweep of both legislative bodies and ALL the other state offices. Governor Pawlenty was the only Republican known to be in trouble early in the election cycle and whose election was known to be really close. Those outstate county ES&S optical scanners were humming away, set up and run for $39 million by the vendor chosen and put in by the former Republican Secretary of State Mary Kiffmeyer and the Republican Governor and the former Republican House.
Are you confident in those results? What good is the paper scanner ballot? Maybe a complex statistical analysis of the historical results of every precinct in the state could find candidates for audit as well as some 'by lot'. Or maybe they could just have used the $44 million to set up a system to count ballots by hand.
List of articles on fixing elections, vote fraud, election manipulation