Nonetheless, the resurgence of the article triggered a lively thought provoking e-debate and introduced again the need to look at elections and counting votes from a business perspective. The importance of the business perspective and business model points of view cannot be overstated. In part because, until we define the problem we're trying to solve, we will never solve the problem. That is to reliably count votes and prove the number of votes we count is correct today, correct tomorrow and tomorrow and tomorrow. One voter --- one vote, one count. Every time. Fraud isn't committed all the time. Mistakes aren't rampant every election. Technological failure doesn't occur every election. But had "we" followed good business practices, we never would have bought into paperless touchscreens (to count 100 votes per machine) as a voting method in the first place. In fact, had we followed good business practices, Florida would have sent 2000 and 2006 back to the voters in precincts where the voting process failed.
From a business model perspective, who owns the ultimate responsibility? Nobody. The fox guards the henhouse. Our election business "model" encourages the classic "he did it / she did it" method to problem definition and solution.
• The Secretary of State certifies the machines as a working and acceptable means of recording and counting votes. Except that as underscored in the GAO's Sarasota reports, this process is blatantly inadequate, in part because the Secretary of State tests ES&S machines following ES&S recommendations using ES&S created test data.
• The Election Supervisor relies on that all important Secretary of State certification to establish a baseline from which to implement his or her ballots for each election. Even the most obscure and seemingly benign faults not detected by the Secretary of State could result in catastrophic failures during an election, particularly one with complex ballots and high turnout. The failure very likely would not be caught by the Election Supervisor's cursory testing to make sure a ballot works.
• When a problem occurs, no one wants to publically admit to flaws in the process. So the voter becomes the convenient scapegoat. The voter has no voice. In effect there is no problem.
Except in Florida, voters' voices were heard. So we rushed out to buy new machines, a new solution. But we still haven't corrected the problem. Our business model has not changed.
Until we develop an election business model that includes high standards for recognizing flawed election results and insists on prompt and consistent corrective action (such as a re-vote by hand if necessary), we cannot ensure all votes will be counted and counted accurately. Until we develop an election business model that requires sound business practices, independent audits, comprehensive computer development and testing methods, and looks rationally at the problem we're trying to solve, we cannot guarantee that any voting technology (DRE, optical ballot scanner, e-voting, mail, or hand counts) will succeed.