Congress should investigate elections

Thursday, January 6, 2005
By KAI STINCHCOMBE   The Stanford Daily

I think Bush won Ohio in this year?s 2004 election. I would even say I?m pretty sure. But given our country?s current electoral system, it?s impossible for anyone to say he?s positive.

First, there are allegations of plain old voter fraud. Precincts in swing states across the country counted more votes than registered voters; voters described getting pre-punched ballots or seeing machines where the ?default? choice was Bush.

In the most egregious potential fraud example in Ohio, Warren County officials cited an unnamed ?security threat? and closed the doors on the press for several hours. It?s impossible to guess what went on in the vote-count headquarters, but studies have suggested that Warren County awarded around 40,000 more votes to Bush than should have been statistically expected.

Many potential-fraud problems revolve around new technologies. Electronic voting machines and tabulating computers for optical-scan votes are vulnerable to hacking, and counts from these machines could be inaccurate. Finally, the hard exit poll data have not yet been released, but statistical analyses based on what was released have suggested likelihoods of less than one in a million of the divergence we saw in this election.

Some of the stuff in the blogs and press is clearly based on statistical errors, but other stuff is credible. When official counts differed from exit polls in Ukraine, State Department officials declared the election fraudulent, but in the United States, the administration says we should ignore the exit polls.

Second, verification was improperly done. Recounts of paper ballots in Ohio were delayed until after the results were officially certified, rather than getting an early start. When the law required manual recounts in randomly ed precincts, the vast majority of counties did not the recounted precincts randomly. This suggests that either they were picking precincts where they knew there were no errors to get through the process easily, or that they avoided recounting votes that they may have known were fraudulent.

In Fairfield County, recount results that didn?t match initial tallies were ?suspended? and begun again from scratch. Press-access rules were once again violated: In Summit County, observers were allowed to supervise recounting but from 20 feet away.

Third, a more subtle campaign systematically depressed Democratic turnout. Machines were moved from Democratic to Republican precincts, so sometimes voters had access to fewer machines than they had used to vote in the primaries. In my old heavily Democratic precinct at Colorado College in Colorado Springs, lines lasted up to 10 hours, while lines in the rest of the heavily Republican El Paso County were under a half-hour. This seems to have been the case at colleges and universities across the country.

Meanwhile, voters were encouraged to use provisional ballots that were not counted rather than find the right place to cast their votes, and different technologies and breakdown rates in different areas meant that poor and black votes were less likely to count than those cast by affluent whites. In the the Bush v. Gore case in 2000, the Supreme Court required states to count all votes in the same way, a sensible recommendation that has yet to be implemented.

All this is just in Ohio.

Our Constitution is based on the principle that only a distrusted government is trustworthy. Concentrated power must be checked by counterpoised authorities, the press and the citizens. While there is no hard evidence of conspiracy or fraud in this year?s election, there is also no evidence of no conspiracy or fraud existing.

The election process could have been designed to produce that sort of evidence, but it wasn?t. Open-source voting machines are more transparent, and voter-verifiable paper trails, like ATM receipts, would mean electronic hackers couldn?t steal an election.

Press-access laws, such as those violated in Ohio, are likewise in place to ensure that officials are accountable to the public.

Finally, independent investigations after elections by apolitical authorities would deter cheating and inspire public confidence in the results. More broadly, steps such as vote-by-mail, same-day voter registration and generous laws about people who vote in the wrong precinct would all demonstrate officials? commitment to make every vote count, regardless of whether it favors their candidate or not.

Today, Congress will meet to accept or challenge the electoral vote results submitted by the states. It would require the vote of one senator in addition to the House representatives who have already pledged to contest the results. It?s possible that our own Sen. Barbara Boxer will lead the charge to open an investigation.

As we all learned early in life, it?s not (just) about whether you win or lose, but how you play the game. I wanted Kerry to win, but the election isn?t just about that. Elections must ensure that everyone gets a fair chance to express their opinion, and that the public is confident in the reported results.

There is almost zero chance, in my estimation, that Bush?s election victory over Kerry was a result of fraud. But that doesn?t mean we shouldn?t investigate. It will inspire public confidence if no fraud occurred; if fraud did occur, we should punish those involved even if they didn?t ultimately affect the outcome of the election.

An investigation would be a good step toward future elections that produce positive, verifiable evidence of the honesty of government officials. That?s the American way.

Kai Stinchcombe is a second-year graduate student in political science and former president of the Stanford Democrats.