From the Christian Science Monitor:

Voting shouldn't require a heroic act of patience
By Alexander S. Belenky and Richard C. Larson
Sep 12, 2006

CAMBRIDGE, MASS. - Of all the issues that decide federal elections, the length of lines at polling stations shouldn't be one of them. Yet, as egregious as that would be, they may. And as implausible as that would be, nobody seems to care. After two straight close presidential elections, the 2006 midterms and the 2008 election are likely to be nail-biters, too. This means that the integrity of the election process matters more than ever.

Is there cause for concern? Yes. Consider the 2000 presidential election. George W. Bush won the presidency by a margin of just 537 votes in Florida. Assume (optimistically) that all the voters were properly registered and marked their preferred candidate, and that no voting machines malfunctioned. Even then, if only 538 Floridians who came to the precincts did not vote due to the widely reported long lines, the election outcome would be in doubt.

In the 2004 election, fewer than 119,000 Ohio votes decided the outcome. Bipartisan accounts suggest that in Columbus, an average of 21 would-be voters per precinct were discouraged by reported waits of four hours or more. If this rate of discouragement held in all 12 of the most populated Ohio counties, with 6,560 precincts - where official tallies show John Kerry won a majority of votes - the result might have been different.

Certainly, long voter lines might have discouraged more Bush voters in each election, narrowing his victories in each case. But we'll never know.

Election queues mostly form when the number of voting machines and support personnel are insufficient to handle swiftly the voters entering the polling station. Culprits include statistical underestimation, incompetence, equipment malfunction, and voter inexperience, especially in dealing with new machines. But deliberate manipulation may also be a factor.

Certain voting precincts can be intentionally "understaffed" with voting machines and personnel. Creating queues can be a potent weapon of partisan election authorities for suppressing votes believed to favor the other party. Among possible abuses that compromise elections, this tactic is difficult to detect, much less to prove. As there are no "exit polls" of voters who gave up because of long lines, red flags aren't raised, and stealth disenfranchisement is a real possibility.

Malfunctions of voting equipment in the 2000 presidential election led to the Help America Vote Act (HAVA), passed by Congress in 2002. But the deployment of voting machines does not have any federal oversight. Two bills that address this issue seem to be bogged down in Congress. Currently, extended election hours are the best that voters can expect from election authorities in overcrowded precincts.

That's why it's time for the federal government to set voting standards, such as a maximum acceptable waiting period to cast a vote, which should be explicitly written in the current bills or in a new bill. Making a maximum waiting period a federal standard would provide "accessibility equity" for all voters.

In the private sector, effective managers of supermarkets, 1-800-number call-centers, and banks employ quantitative analysis to keep customer lines moving. In professional hands, the tools of queuing science and optimization are sufficient to meet the challenge. In elections, the same tools should be the basis for a new deployment system for fairly allocating voting machines and personnel across precincts.

Supplemented by Election Day "emergency measures," such as an adjustable redeployment of mobile voting units, this new system would greatly decrease the numbers of voters lost to unexpectedly crowded precincts. Unpredicted crowds, of course, could eventually cause voter queues even if the system's recommendations are correct. However, such a system would make deliberate manipulation to suppress the turnout of particular voters much more difficult.

The integrity of the election process is severely damaged when long lines keep voters from casting ballots. By pressing Congress to establish strong voting standards, we, the voters, can help preserve the integrity of American democracy.

* Alexander S. Belenky is a visiting scholar at MIT's Center for Engineering Systems Fundamentals. His third book on US presidential elections, "How America Chooses Its Presidents," is forthcoming. Richard C. Larson, director of the center and MIT's Mitsui professor of engineering systems, has written many scientific articles on queuing theory.