HAVA Look: A Simple, Verifiable, Open Source, Paper Ballot Vote-Recording & Counting System -- TIA

TruthIsAll     http://www.geocities.com/electionmodel/HAVAOnePC.htm

Thanks to HAVA, elections are vulnerable to multiple vote miscounts: at the precinct level with closed-source, diverse, proprietary code-based touch screens (DREs) and Optical scanners and then on closed-source, proprietary Central Tabulators. They have you coming and going.

On the other hand, paper ballots can easily be input to a standard personal computer spreadsheet. Voters can request a copy before leaving the precinct. The precinct (spreadsheet) file is uploaded to the Internet. The voter can quickly and easily verify his vote as the spreadsheet ballot records are sorted by a unique Precinct/Ballot ID code. At the same time, he can view and/or download the total precinct ballots to verify the vote counts. Consolidation of district/county/state vote counts is tabulated by Open Source software.

Three (3) volunteers input and verify the paper ballots. As Volunteer A reads the ballot, B enters the data and C verifies the result. Data entry would be completed shortly after the polls closed. The voter fills out a ballot marking each selected candidate's code (i.e. a=Dem, b=Rep, c=Ind). The volunteers use their god-given optical scanners (EYES) to read the ballots and enter the codes in a spreadsheet. As each code is entered, the spreadsheet automatically calculates the total vote and percentage for each candidate.

The following simple steps complete the process:

There are approximately 200,000 precincts nationwide, an average of 4,000 per state. How much would it cost and how long would it take to process and count the votes in an average precinct of 600 voters?

Just one (1) $400 Personal Computer is used for ballot data entry

$100 Printer

$100 Spreadsheet

Assuming that data entry/verification takes 1 minute per voter, total data entry time would be 600 minutes (10 hours), which can be spread out over 12 hours on Election Day. Larger precincts could install additional PCs and add volunteers as necessary. The individual PC spreadsheets would be combined into a single file (sorted by Ballot ID) for uploading to the Internet.

There are many advantages over the current, vulnerable system. HAVA look:

• Data redundancy:


1. Original paper ballots: retained in a secure location
   
2. Precinct spreadsheet files: ballots sorted by ballot ID
   
3. Internet: copies of uploaded precinct files for online viewing/download
   
4. Internet: County/state composite files (vote totals sorted by precinct/county) for online viewing


• Data / software integrity:


1. Original paper ballots: documented chain of custody
   
2. Data integrity (100% automatic audit to catch discrepancies between the local precinct spreadsheet and corresponding internet file
   
3. Open Source code: eliminate proprietary code on DRE/ Optiscan/ central tabulator


• Eliminate corrupting influence of vendors and election officials:


1. Hardware: Eliminate bribery of election officials for purchase of DREs and optical scanners
   
2. Diebold/Premier & ES&S out of the DRE/Opscan voting machine business

A Simple Precinct Spreadsheet Vote Recorder

Example:
Precinct 101 (500 voters)
Party Code: a=Dem, b=Rep, c=Ind
Voter Ballot Code: 101xxxxxx

Note: Totals are calculated using the Excel function: COUNTIF (range, party code).
For example, the total number of Democratic (column B, rows 10-509) votes for president:
242 =COUNTIF (b10.b509,”a”)

            Pres   Senate  Congress  Other1   Other2   Other3   Other4   Other5   Other6   Other7   Other8
Totals
Dem          242      233      252      242      233      252      242      233      252      242      233
Rep          234      244      225      234      244      225      234      244      225      234      244
Ind           24       23       23       24       23       23       24       23       23       24       23
BallotID
1010001        a        a        a        a        a        a        a        a        a        a        a
1010002        a        a        a        a        a        a        a        a        a        a        a
1010003        a        a        a        a        a        a        a        a        a        a        a
1010004        b        b        b        b        b        b        b        b        b        b        b
1010005        b        b        b        b        b        b        b        b        b        b        b
1010006        a        a        a        a        a        a        a        a        a        a        a
1010007        a        a        a        a        a        a        a        a        a        a        a
1010008        b        b        b        b        b        b        b        b        b        b        b
1010009        b        b        b        b        b        b        b        b        b        b        b
1010010        b        b        b        b        b        b        b        b        b        b        b
1010011        b        b        b        b        b        b        b        b        b        b        b
... ...
1010490        b        b        b        b        b        b        b        b        b        b        b
1010491        b        b        b        b        b        b        b        b        b        b        b
1010492        b        b        b        b        b        b        b        b        b        b        b
1010493        a        a        a        a        a        a        a        a        a        a        a
1010494        b        b        b        b        b        b        b        b        b        b        b
1010495        a        a        a        a        a        a        a        a        a        a        a
1010496        b        b        b        b        b        b        b        b        b        b        b
1010497        a        a        a        a        a        a        a        a        a        a        a
1010498        a        a        a        a        a        a        a        a        a        a        a
1010499        a        a        a        a        a        a        a        a        a        a        a
1010500        b        b        b        b        b        b        b        b        b        b        b	

The OP describes only a Subsystem of an entire Voting System, namely one which deals with VERIFYING the vote by 1) offering a line-item report of the vote for scrutiny by HUNDREDS of VOTERS' eyes to OPENLY verify the recording of the vote by the select few humans delegated to handle the PAPER BALLOTS, with a "side effect" of 2) affording a count of the vote that is indepdendent of the select few humans delegated to count the DETAILS on the PAPER BALLOTS

It has NOTHING to do with DREs or Optical Scanners, OR THE PROBLEMS OF SECRECY/PROPRIETARY CODE/UNRECOUNTABILITY etc.

Diva's sentences:
"logistically, the machines are NEVER all examined prior to an election;"
"they are distributed weeks or days in advance of the election, they are subject to the same security failures and breakdowns as proprietary devices;"

The difference is a "standard PC" -- if defective -- can be easily and quickly replaced, likely in much less than an hour! They are cheap enough to have a backup onhand. Moreover, the OP is talking about a PAPER BALLOT environment!

READ THE TITLE: "A S..., Ver..., Op..., PAPER BALLOT ... ..."

"Paper Ballot" means a physical, analog record of the vote exists, and, thus, the absence of a working PC and/or software would not preclude conducting any standard HCPB procedure. The presence, though, of a working PC and software can *facilitate* and greatly enhance the integrity of any PAPER BALLOT election, at and beyond the precinct level. (See "advantages" in the OP)

The only "machine" involved is a standard PC, which -- in the CONTEXT OF ITS USE in the OP -- DOES NOT HAVE the "SAME" security risks as other "voting" "machines" (and should NOT be so relegated for purposes of estimation), since

• It's not involved in any secret recording the vote (like a DRE) -- HUMANS openly read and record data from PAPER BALLOTS onto the spreadsheet & preserve the paper ballots
  
• It's not involved in any secret counting of the vote (DRE, OpSc) -- HUMANS can do the count (but Excel can record, count and produce organized output for efficient, independent scrutiny of its input data)
  
• It's not involved in any unverifiable counting of the vote (DRE, OS-if-the-"audit"-is-half-assed)
  
• It produces practical physical output at the precinct and higher-up levels, for verification of the vote at every level by HUNDREDS of EYES (unlike DREs, OpSc, and HCPB-methods) ...AND I GUARANTEE MORE THAN 3% of VOTERS will be "AUDITING"
  
• There is no code hiding secret manipulations of data (like on DREs or OpSc), since each analog-recorded data entry to spreadsheet is verified by humans (...and, in the context of spreadsheet usage, RE-VERIFIABLE by voters themselves, not only at the PRECINCT but also via INTERNET and also RE-RE-VERIFIABLE AFTER the vote's upload to Central tabulator on Election day itself).
  
• If a standard PC "breaks down", it is easily, quickly replaceable at local stores, likewise "off-the-shelf" software like Excel. ANY and EVERYTHING INPUT and OUTPUT via WHATEVER PC & Software USED is OPEN to INSPECTION and CROSS-CHECKING at EVERY LEVEL by NUMEROUS PEOPLE, including voters themselves (who, also, become empowered to ADD UP THE COUNT themselves). Can always resort to standard HCPB methods (whatever those are)
  
• The spreadsheet is primarily organizing output for verifying the vote ON THE DAY OF THE ELECTION. Since a COUNT and RE-COUNT of the Vote CAN ALWAYS proceed from the Paper Ballots themselves by hand tallies, it need NEVER be deemed an unwelcome situation that Software and the EYES OF HUNDREDS OF VOTERS can provide indepedendent reviews and counts of the data as a check against human error, whether in recording, counting or both.
  
• What the human handling and hand tallying cannot produce is a neat line item report for EVERY G**DAMN voter to VERIFY that HIS OWN G**DAMN PAPER BALLOT VOTE had been properly recorded (by parties likely unknown to him, by whatever method of recording used).
  
  "sorry, it's just putting lipstick on a pig in my humble opinion to think that open source is going to be a viable alternative. Machines be damned!!"
  
  uh huh...an expected cliche, but a crude over-reach here...depends on the context and the details.

For the presumably numerous parties who might be privileged participation in an Election Day hand count of Paper Ballots -- i.e. forget Levers (virtues and drawbacks), DREs, and optical scanners...the OP deals only with voting systems based on auditable, individual Paper Records of voters' intent.

These ballots have neither 1) been recorded in secret (open to deliberate vote-switching) nor 2) are to be counted in secret by any non-transparent, intermediary electro-mechanical device (open to deliberate mis-counting). The unavailability of line-itemized printed output would be a missed opportunity for not only verification of the vote, but also protection against vote-switching and non-counting that can occur at the precinct level and beyond.

The fact that physical ballots in Ohio have been transported between precinct counting machines and have eventually been "destroyed or lost" altogether, attests to the extent that election fraudsters have gone to systematically undermine democracy in America and to the vulnerability of paper records.

The inclusion of a simple precinct spreadsheet procedure need not be an exclusive proposition: It can be done in parallel with whatever other HCPB methods of counting might be utilized and would be non-obtrusive. The benefits in preserving a verified "backup" of the paper records themselves should be considered.

Vote Secrecy can be protected with proper procedure and method. It is presumptuous to assert voter-privacy need be compromised.

Comments and Criticisms

The author has had over 30 years experience in quantitative application program design and development—from defense/aerospace IBM mainframes to Wall Street PCs. He converted main frame applications for investment banks on Wall Street from FORTRAN to Lotus 1-2-3 in the early eighties and then to Excel in the mid-nineties. He developed and marketed spreadsheet based corporate financial models to some of the largest U.S. consumer product manufacturers and foreign banks. As a programming consultant, he developed Excel VBA applications for dozens of the world's largest international banking institutions and corporations. He knows very well what Excel is capable of.

Critics created an Excel straw man which reveals limited knowledge of programming in general and spreadsheets in particular. They appear to have an agenda to retain voting machines which can be rigged internally and/or by humans after the fact. The post suggests an Open Source system developed by professional programmers for use in all state precincts — not a set of individual spreadsheets written by non-professionals. A robust Excel VBA application could be developed in a matter of weeks.

In response to a challenge to corrupt the spreadsheet, one critic, a database proponent, said: "No thanks". That says it all. Why not take up the challenge? And not one word about the Access software for voting machines which has been proved vulnerable to hackers time and again.

It is patently false to claim that spreadsheets are not up to the task (due to security issues) and that a database system is the only feasible programming solution. Virtually all banking institutions use Excel in conjunction with securities databases for data retrieval, input and analysis. Criticism of spreadsheets is reminiscent of that leveled at Lotus 1-2-3 and PCs in general by MIS Luddites in the early eighties. Corporate gatekeepers feared that mainframes were being made obsolete by the new machines. They called them toys. Some toys. Spreadsheets have long been used for securities data retrieval and analysis by all U.S. corporations — from Wall Street Investment Banks, consumer goods manufacturers, consulting and accounting firms et al.

Have critics ever written a commercial Excel-based application? One can only assume from the comments that they haven't. Then why don’t they create an equivalent program using their favorite database? That should be a piece of cake for them. The sample spreadsheet took all of 20 minutes to write and proved that counting votes is not rocket science.

Critics fail to suggest a new system design to help eliminate election fraud. What hardware would they use? What software? Would they keep DREs and Optiscans? Should precinct records should be available online for individual retrieval? If yes, how would they go about it? If not, why not? Diebold used Microsoft Access software. How did that go?

The Coopers & Lybrand "study" cited in the article is highly critical of spreadsheets. The study was probably done twenty years ago: "Coopers and Lybrand in London cited research showing that over ninety percent of all spreadsheets with more than 150 rows contained at least one significant formula mistake". But there is little doubt that C&L still uses Excel.

This statement is a complete falsehood which exposes the author's naivete: “Spreadsheets are often temporary pieces of work. Generally, they are used to meet a particular need and then discarded, or used very little. This means that the full rigors of software engineering are often absent from their development. The ease with which an inexperienced user can produce plausible output also tends to militate against the adoption of a more rigorous approach Ray Panko of the University of Hawaii has collected information from both field audits and laboratory experiments in spreadsheet development indicate that spreadsheet errors are fairly common.”

Is the author stating that spreadsheets are inherently error-prone? That is preposterous. Were the "errors" cited the result of sloppy coding by individuals who were office workers as opposed to professional programmers? Were they the result of data entry errors? This statement shows the writer’s ignorance, inverted logic and misplaced sarcasm: "In another study, subjects were shown large and small, well–formatted and poorly–formatted spreadsheets. Which combination inspired the most user confidence? You guessed it: large, well–formatted ones. Call it the Information Age Effect; with all that rigorous data so beautifully laid out under program control, how can anything possibly be wrong"? What the author doesn't realize is that a well-formatted spreadsheet, just like easy to read programming code, is a key element of good program design. Mistakes are more likely to occur when a program or spreadsheet is convoluted and complex. This applies to all current programming environments: C++, Visual Basic, Java, etc, as well as the original mainframe procedural language compilers (FORTRAN, PL/I, COBOL).

The author's objections to Excel are the equivalent of GOP/media talking points—just like WMD and terror alerts. No substance. He cites these Excel "disadvantages": Repeated Data; Data Entry; Data Validation & Checking; Data sharing/ Collaboration; Search and Retrieval; Error Debugging; Upper row limit (65000 for Excel); Easy data duplication; Formula errors. But he does not support the statements.

He has no clue about Excel's powerful facilities and is obviously unaware of these relevant features (see Excel Help for examples and documentation):

• Object-oriented programming: Excel Object Model, Visual Basic, interface to programming languages such as C++.
  
• Databases: Data objects (DAO, ADO); Jet SQL
  
• Security: virus protection, digital signatures and certificates; workbook, worksheet and cell protection
  
• Web interface: files, hyperlinks, archiving, web servers
  
• Debugging tools: available for the programmer to check VBA code, data and formulas.

Every programming language requires that the programmer step through the code. Excel is no different. Formula and programming errors are caused by programmers during development. These errors are not unique to spreadsheets; they are unique to humans. Sixty-five thousand rows are more than enough to handle the largest precinct in the U.S.  Excel critics are unfamiliar with Visual Basic for Applications (VBA). Have they had hands-on commercial and/or scientific programming experience in mission-critical applications?

How would critics insure the integrity of the vote? Anyone who has done any reading on the election fraud would agree that there no way to PROVE that votes are counted correctly. Here's a brief summary of why all voting methods can be rigged:

DRE touch screens: Insecure
Many hardware/software loopholes for vote-switching have been exposed

Optical scanners: Vulnerable
No chain of custody; limited or no recount allowed

Levers: Unverifiable
Vote totals reversing at 99, zero votes for Obama in Harlem

Punched cards:
Pre-election under-votes; post-election over-votes

• In Florida 2000, 110,000 double and triple-punched over-votes, 75,000 under-votes; Dan Rather produced an explosive documentary on how faulty paper was used in 2000 to limit the vote count in heavily Democratic Palm Beach County.

Central tabulators: Invisible
Who controls the count controls the votes. Stalin would be proud.