Or, how to figure out when bribery and threats sway elections.
By Nathan Collins
Nicolás Maduro clenches his fist after he was sworn in on March 8, 2013. (Photo: Juan Barreto/AFP/Getty Images)
In the popular imagination, election fraud usually takes a few forms: stuffed or disappearingballot boxes, hijacked electronic voting machines, voter intimidation. The latter of those four tactics is somewhat harder to detect. After all, it’s not the ballot being tampered with, but rather the voter who cast that ballot.
Fortunately, researchers have now figured out a way to detect “voter rigging,” as the authors of a new paper call it. Unfortunately, their method has turned up more or less exactly whatyou’d expect—fraud in Russia several times over the past decade, as well as in Venezuela, where voter rigging likely swayed the outcome of the 2013 race to replace Hugo Chavez.
“Many elections around the world end in controversies related to alleged frauds; even in mature democracies, such as the U.S. and Canada, where voter suppression scandals have made the headlines,” write Raúl Jiménez, Manuel Hidalgo, and Peter Klimek.
Of potentially greater concern: detecting fraud in countries that have the trappings of democracy while maintaining at least some degree of authoritarian rule.
Of potentially greater concern to the researchers: detecting fraud in countries that have the trappings of democracy while maintaining at least some degree of authoritarian rule. “For example, there are viral videos filmed during Zimbabwe elections that show electors that were forced to vote or allegedly bused under intimidation,” the researchers write.
The question, then, is how to detect that kind of thing without the benefit of videos and other direct evidence? “Our key hypothesis is that small polling stations are more susceptible to voter rigging, because it is easier to identify opposing individuals, there are less eye witnesses, and supposedly less visits from election observers,” Jiménez, Hidalgo, and Klimek write.
If that’s correct, then fraud would show up as a spike in turnout or the ruling party’s vote share in small towns, but not in large cities. (In particular, neighboring large cities. While those won’t necessarily vote the same as small towns, differences between neighboring municipalities ought to at least be fairly stable over time.) For example, if a small town’s voter turnout is higher than normal and the vote goes to the ruling party, and something similar isn’t happening in large cities nearby, that could be a sign of fraud.
Using that approach, the researchers identified a number of elections in which overall turnout and the winner’s share of the vote was unusually large in small but not large polling areas, suggesting fraud, particularly in Russia and Venezuela. “In the Venezuelan case, where the current president (Nicolás Maduro) was elected by a plurality voting system, the systematic distortion of voting behavior in small electoral units was outcome-determinative,” the researchers write.
Not that their method can detect all cases of voter rigging. As they point out, if bribery and extortion are widespread—as they are thought to have been in the 2011 Ugandan elections—the new approach may fail to detect fraud. “[I]t is unclear to us which of the following two facts is more disturbing,” the researchers write. “[N]amely (i) that such large-scale distortions of vote preferences keep recurring to the point that they may be outcome-determinative or (ii) that such practices are so blatantly committed that they are hidden in plain statistical sight.”