There was little in the way of transparency in the Iowa central count that produced the final result. The Republican hierarchy chose to conduct the count in total secrecy and, possibly, out of state in Kansas. Party officials said the secrecy and mystery location resulted from fears that some Occupy movement would disrupt their counting.
Fortunately for citizens, Watch the Vote 2012, a voting rights activist group, was present to monitor the caucuses. They had a presence in approximately half of Iowa's counties.
That allowed a check of the relationship between precinct size and vote accumulation for each candidate. The records also allowed a check of precinct totals against those produced by the secret tabulation of precinct totals.
The precinct voting by activists did not match up with the party data presented. Edward L. True, a caucus participant, was wise enough to photograph his caucus count. It was different from the party count; the discrepancy resulted in a recount that changed the outcome from a Romney to a Santorum victory. The finding was too late to do any good but it made an important point.
The New Hampshire primary was also monitored by the WatchTheVote2012 activist group. They monitored hand counts at voting places. This allowed a comparison of hand counts, created without any computerized intervention, with the count of the central tabulator for the state provided by a private company, LHS Associates.
The hand count precincts showed no amazing anomaly, while the central tabulator producing totals for precincts revealed the amazing anomaly, progressive and highly improbable vote gains for Romney as precinct size increased.
The results from these two states demonstrate that central tabulation does not necessarily require computer involvement to produce highly questionable results . In this case, questions persist due to secrecy and the potential for human interference. In the case of New Hampshire, citizen diligence and involvement made the difference in detecting irreconcilable differences between no-tabulator results and results produced by a central tabulator that appear to make no sense. .
Are Central Tabulators Vulnerable to Hacking and Manipulation?
The best way to hack a central tabulator is to buy an elections equipment company. These privately held, unaccountable firms provide the majority of voting systems in the United States. ES&S acquired Premier, successor to the ill-reputed Diebold franchise, but was forced to sell it to Dominion. These two plus Hart Intercivic dominate the market for elections systems, from voting machines to central tabulators to service contracts for full election management.
What better place to control an election outcome, tailor made for whomever.
These firms sell or otherwise provide equipment with the provision that the software behind the machines is their sole intellectual property. County and state election officials are unable to inspect the software at multiple levels due to this restriction. This is truly secret voting outsourced to private firms who operate behind the scenes and have limited accountability. There is no true public accountability.
In his Ars Technica article on the subject, "Miscounting the Vote," Jon Stokes outlined methods of attacking a central tabulator. These apply today. Blackboxvoting.org has demonstrated the vulnerabilities to system hacks, even showing how a trained chimp could perform the task.
Francois suggested his own variations on hacking a central tabulator in an effort to update Jon Stokes.
"Method #1: Using an optical scan voting form, DRE, or absentee ballot, encode the form or use the keyboard to issue commands to the central tabulator (CT). Therefore the VOTE ITSELF is used to hack the CT. For example: the Riverside 2012 General election includes 20 elections. Fill out the ballot for each of the first 10 elections as follows: 1,2,1,2,1,2,1,2,1,2 This "wakes up the CT'. Then fill out "3, 1, 4, 2, 2, 1, "' which commands the CT as follows: "in election #3 flip candidate #1 up; in election #4 flip candidate #2 up; in election #2 flip candidate #1 up' and so forth.
"From that point, the CT software knows exactly what to do. It won't bother with candidates with less than 10% of the vote. But, it will flip progressively as a function of precinct size and try to flip just enough to win, etc. By communicating commands through votes, NO INTERNET to the CT connection is required. No local accomplices at the County level are required. The "fix' could have been programmed in the CT software as far back as decade ago, with no knowledge of this year's political candidates.