30 May 2008
Election Fraud Difficult to Detect, Measure
Expert panel discusses fraud and other obstacles to free and fair elections
Washington -- Especially since the close 2000 U.S. presidential election in which Americans and the world watched poll workers count card-punched ballots by hand, there has been growing interest in detecting and preventing election fraud.
But election problems rarely are the result of fraud, experts suggested at a panel hosted by the Brookings Institution in Washington May 21. The experts, authors of the book Election Fraud: Detecting and Deterring Electoral Manipulation, discussed how to detect and measure the effect of electoral fraud.
For an election to be considered fair, individuals who are qualified to vote must be allowed to vote, votes must be counted fairly, unqualified votes should not be counted, votes should not be coerced and voters should not be rewarded or bribed for voting.
Thad Hall, assistant professor of political science at the University of Utah, added that a fair election allows a voter to choose not to vote. "It's something we don't normally think about, but you should be able to freely disengage in the process."
An election can be unfair and still not involve fraud, the experts said. Fraud has to be an intentional act, the authors said, and that can be very difficult to prove.
"There is no perfect election, ever. There is always some flaw somewhere," said Susan Hyde, assistant professor of political science at Yale University.
ELECTION FRAUD RARE, BUT EVEN PERCEPTION OF FRAUD IS DAMAGING
R. Michael Alvarez, professor of political science at the California Institute of Technology, said that since 2000, he and Hall "have probably observed … hundreds, maybe thousands, of polling places, and we have never seen any evidence of fraud."
"But we have seen a lot of evidence of a lot of other problems," Alvarez said. Some of these problems are "probably affecting the outcome of elections in a much more sweeping way than actual fraud."
"Perceptions of election fraud, even if they are not accurate, can be just as destructive to an electoral process as actual election fraud," said Hyde.
"The context of what is considered fraud varies by where you are and the country you are in," said Hall. For example, some countries may consider it fraudulent for a current president to endorse a presidential candidate. Others may consider door-to-door campaigning a pressure tactic. Both activities routinely occur in American politics.
The definition of fraud also changes over time. Decades ago, the United States did not allow large segments of the population, including women and African Americans, to vote. Today, purposefully disenfranchising these groups would be fraud.
Election fraud can occur at times other than Election Day. "Election manipulation can be conducted before the campaign starts, during the campaign period, on Election Day and during and after the tabulation of results," Hyde said. For example, a person can falsify petition signatures to enable a candidate access on a ballot. A person also can threaten voters, or prevent leaders of the political opposition from running for office.
Measuring fraud is extremely difficult. With secret ballots, explained Hall, "I cannot know how any person in this room votes. … If I could though, I could make sure that there was no fraud."
However, experts can use data to understand traditional voting patterns and note unusual incidents that deserve more careful scrutiny for potential fraud. To do this, experts need access to large amounts of data and need election officials to be transparent about the process. This can be a challenge in other countries as well as in the United States, where some states do not allow election monitoring, said Hall.
To measure fraud accurately, the authors said, experts need to work with election officials to help them understand that there are easy ways to be transparent and provide useful data. Countries such as Estonia, Argentina and Indonesia have provided good examples of how to do just that, the experts said.
By releasing precinct data, allowing election monitors and letting federal officials meet with political experts, these countries have shown that it is, "not so difficult to be transparent," said Hyde.
