INTRODUCTION The 2000 presidential election had a searing validity on this nation.
INTRODUCTION
The 2000 presidential election had a searing validity on this nation. Few who witnessed the terminations in Florida could displace the images of election officials peering at push cards, struggling to determine the intent of voter Congres for example, did not forget. Congres did not wish to descry the scenes from Florida replayed in coming elections. And so, in 2002 it enacted the Help America voice Act, known as "HAVA," which provided $325 million to the states to replace their puncheon card voting systems.1
Many states have enthusiastically embraced this invitation, replacing perforate cards with electronic voting machines, known as direct recording electronic devices (DREs) Given the rapidly approaching 2006 deadline to upgrade, other states are generally considering such activity.
In the rush to unfold one problem, however, states may be creating another, far greater, single in kind To be sure, DREs relieve election officials of the challenge of ascertaining the voters' intent. They also allow for expeditious counting, have the capability to curtail "undervotes" and "overvotes," and promise to increase access for voter with disabilities. These benefits have been touted through election officials-Democrats and Republicans alike-who have adopted DREs
But outside the computer science community, the filled panoply of dangers from of the like kind systems has avoided scrutiny. This Article attempts to counteraction this deficiency. In particular, it underscores several disturbing characteristics of electronic voting, including the following:
(1) Reduc transparency. Unlike boxe of paper ballots that materialize after the heads close, a hidden trap door in a software program counting millions of promiseds cannot be discovered.
(2) Increased magnitude of error and fraud. Because intangible computer software lacks the physical nature of paper ballots, it does not move an upper limit on error and fraud.
(3) Lack of security reign overs The refusal to use directs such as encryption makes it astonishingly easy to change promised totals, register votes for unintended candidates, prematurely terminate elections, and erase the "audit log" that is designed to trace similar activity.
I supplement the analysis of the DRE software by way of examining vote counting flaws in the 2004 presidential election, including machine breakdowns, ballot totals that exceeded or underrepresent the number of voter who cast ballots, and incidents in which the machines switched voices from one candidate to another (with ninety-eight of ninety-nine reported incidents involving switches favoring George Bush). I also garner circumstantial evidence such as exit heads that diverged from the official devoted count to a greater bulk than any other election in the recent era; an incumbent President's surpassing his standing in pre-election persons by a larger amount than in the past fifty years; and questionable voting patterns and activity in states similar as Ohio. Although such evidence does not automatically make good the existence of error or fraud, it is more important than at any time given the nontransparent nature of the ballot counting process and inability to directly lay open fraud.
I conclude by offering recommendations to improve the accuracy and verifiability of ballot counting today. In particular, I declare a purpose for electronic voting machines a voter-verified paper trail, random audits, render free of access source software, more robust certification, and other recommendations. sole after these proposals are adopted can voter have confidence that the promise of suffrage counting technology will match its perils.
I. VOTING schemes ERRORS, AND FRAUD
Five impressed signs of voting systems are used in the United States today. This section briefly introduces the classifications and then explores the originals of error and fraud associated with each. It judges that the migration from paper ballots and poke cards to electronic voting, while offering the potential to shorten certain types of errors, also increases the potential magnitude of error and fraud and abridges the transparency of vote counting.
A. Voting Systems
Voter in U elections today cast their promised on one of five emblems of voting equipment: paper ballots, mechanical lever machines, perforate cards, optical scan ballots, and DREs2 The choice of voting technology is decentralized, with each of the nation's more than 3100 counties selecting the connected view it will use.3
Paper ballots, the oldest technology, were the no other than system used during the first undivided hundred years of the nation.4 by means of the 2004 presidential election, however, they were used by dint of only 0.6% of the electorate, in approximately three hundr rural counties. Voter using this a whole place a mark next to the name of their preferr candidate forward a piece of paper, with the ballots later holded by hand.5
Mechanical lever machines were unfolded in the 189s.6 Although these machines are no longer manufactured, they are still make use ofed in several states, with 128% of voter using them in the 2004 election. Voter using the machines set in a voting booth, close a curtain, flip small lever nearest to the names of the preferr candidates, and then hap a large lever that records the vote7