Interim Voting System: Automated One Vote Per Person Verification By George Edwards
Before the Most Recommended Voting System--Automated Fingerprint Verification is in place, an interim voter card management system to automate one-vote-per-person verification could be implemented in anticipation of later conversion to the most recommended system for which this is a subset. This interim system could provide added protection against voter fraud than some current systems.
In a nutshell, a unique number is associated with every voter card, the number is automatically compared against previous voter card numbers used in a precinct in a particular voting cycle. If there is a match, poll workers are alerted. If the voter, says that he or she has not voted earlier, the poll worker can allow the vote to be entered, but should alert the voter that a counterfeit card has apparently been used and ask to pick up the card used by the voter, promising that the voter will be sent a new card with a different bar code. If the old bar code number is later used, this would necessarily be fraudulent and a voter attempting to use it should be prosecuted. For maximum protection, the card number should be unique country wide with a portion uniquely identifying the state and the precinct.
In South Carolina, for instance, there will be a bar code on the to-be state issued voter photo ID card. So I gather there is some thought directed towards a bar code reader at the voting place. I have no idea what further is already in mind. This bar code approach could work for an interim voter card management system anywhere.
In any case encryption/decryption techniques recommended by experienced system or card management systems [CMS] vendors should also be used.
The voter list of those registered to vote in a precinct, properly signed and monitored, provides protection against multiple votes by the same voter in a precinct. It does not protect against those improperly registered as to living in other precincts or states or against lax performance at a precinct.
The most recommended system includes provisions to automatically provide protection against multiple votes cast by the same voter in other jurisdictions than a particular precinct. Similar provisions could be used in an interim system as an aid to alert poll workers of multiple votes by individuals in other jurisdictions. And having the foresight to incorporate them now would also minimize the needs for upgrading to the most recommended voting system—limiting the needed upgrade to the fingerprint verification it would provide via smart cards.
As a minimum in an interim system, a unique number on each card for every voter in a precinct would be used as a way to help further guard against multiple votes by the same voter if the number is compared with such numbers stored in precinct-machine memory for previous votes in that election.
This, in addition to a state and precinct number, is included in the “Most Recommended Voting System--Automated Fingerprint Verification.” So if such provisions were made in putting together a bar code, or other, card* reader system to alert poll workers as to the authenticity, or not, of a vote, this would already be in place if and whenever an automated fingerprint verification system might be implemented.
The most recommended automated system will not record such an automatically detected duplicate vote without operator intervention. In the interim system, poll workers would be automatically alerted when a duplicate vote has been detected. If the voter, says that he or she has not voted earlier, the poll worker can allow the vote to be entered, but should alert the voter that a counterfeit card has apparently been used and ask to pick up the card used by the voter, promising that the voter will be sent a new card with a different bar code.
If the old bar code number is later used, this would necessarily be fraudulent and a voter attempting to use it should be prosecuted.
Changes already have to be made to voting-machine software in every election cycle to correlate the names of new candidates with the precincts. In the most recommended system, to accommodate voters who have changed precincts since last voting without requiring additional provisional ballots or new cards, the former precinct number would also be newly correlated in the voting machine software with that set of new candidates. It makes good sense to do the same in an interim system as well.
In the most recommended system, the field for complete unique fingerprint numbers provided to prevent multiple votes anywhere in the nation would have to be large enough to accommodate numbers associated with all current and anticipated voters nationwide for some extended period, say a century, unless deceased voter-number vote-counting is removed from the system as soon as possible after deaths are reported. Again, it makes good sense to make the same provision in the interim system. With an automated fingerprint verification system, voting of the dead would not be an issue.
Smart cards are necessarily used in the most recommended voting system to accommodate the fingerprint data. Their use in an interim system would likely be an unwise expense, but the inclusion of provisions for the computer-aided defense against multiple votes by the same voter is worthy of consideration for the interim system itself as well as to minimize the conversion requirements to the overall most recommended fingerprint-verification-voting system.
*Quoting from http://en.wikipedia.org/wiki/Card_reader#Smart_card_readers:
“The advantage of using barcode technology is that it is cheap and easy to generate the credential and it can easily be applied to cards or other items. However the same affordability and simplicity makes the technology susceptible to fraud, because fake barcodes can also be created cheaply and easily, for example by photocopying real ones. One attempt to reduce fraud is to print the barcode using carbon-based ink, and then cover the bar code with a dark red overlay. The barcode can then be read with an optical reader tuned to the infrared spectrum, but can not easily be copied by a copy machine. This does not address the ease with which barcode numbers can be generated from a computer using almost any printer.”
Of interest, this same reference discusses the pros and cons of other types of cards, specifically, biometric, magnetic stripe, Weigand, proximity and smart cards.
As to, so-called smart cards, to quote from this reference: “Card memory may be used for storing biometric data (i.e. fingerprint template) of a user. In such case a biometric reader first reads the template on the card and then compares it to the finger (hand, eye, etc.) presented by the user. In this way biometric data of users does not have to be distributed and stored in the memory of controllers or readers, which simplifies the system and reduces memory requirements.” [And, in my mind perhaps most importantly, removes the need to access a remote data base, with the added security problems that imposes.]
Continuing: “Smartcard readers have been targeted successfully by criminals in what is termed a supply chain attack, in which the readers are tampered with during manufacture or in the supply chain before delivery. The rogue devices capture customers' card details before transmitting them to criminals.”