Instant Runoff Voting as Addressed by Dr. Rebecca Mercuri. April 10, 2007
Dr. Rebecca Mercuri, whose testimony to the NC State Legislature was influential to passing our Public Confidence in Elections Law - has serious concerns about IRV. With her permission I post them here:
"IRV and other proportional balloting methods have been proven to incentivize the introduction of electronic ballot tabulation in places where none previously was needed or has existed, and they further complicate whathas become an increasingly closed process for the determination of election results.
For example, in May, Scotland will see their first use of a run-off style election, accompanied by their first use of optical ballot scanners. Here in the USA, where voting system vendors are allowed to sell products that are encumbered by trade secrecy such that the examination of their tabulation software is precluded, there is a growing list of instances where straight-forward vote calculations have been discovered to have been mis-programmed, following the election. One of the most horrendous of these caused the tabulator to begin counting backwards (deducting votes) when a numeric overflow was reached.
Other such mistakes have involved switching votes from one candidate or party to another. There is no way to determine the extent that undetected programming mistakes have affected election results, because audits have not been required and many of the voting systems lack any way to perform an independent audit.
Since the program code that controls the voting system cannot be reviewed, election officials have had to resort to the use of logic and accuracy testing prior to each election, in order to attempt to ascertain whether or not the balloting and tabulation equipment is functioning properly. Unfortunately, this testing typically cannot exercise all of the selection combinations that are possible, on all machines, for each election, even using the traditional vote choice method. The problem of creating a sufficiently large and varied test ballot set that is capable of
revealing whether the calculations have been programmed properly for the election, is further exacerbated when compounded with run-off methods of voting.
It is incumbent upon election officials to be able to thoroughly understand how results will be calculated, such that manual recounts and audits are possible, when mandated. These processes also become increasingly difficult and more time consuming when run-off methods are employed, so the chance that errors in the electronic or manual vote calculations will be detected and properly resolved is necessarily reduced.
Furthermore, there are certain run-off methods that can produce different results based on the organization of ballots in the stack.
Since these methods lend themselves to potential "gaming" of the ballot set that may not be independently detectable or auditable, these run-off styles must be prohibited. The present climate of distrust regarding election integrity will only be further undermined by skepticism invoked by increased complexity of alternative balloting methods, especially if the vendors are allowed to continue to obfuscate their vote tabulation products."www.notablesoftware.com
http://www.ncvoter.net/downloads/Dr_Rebecca_Mercuri_Instant_Runoff_Voting.pdf