The Iowa City Press-Citizen published the following editorial on June 8, 2010, primary election day in Iowa:
Studies by political scientists show party identification stands out as the single most important factor by which voters choose between candidates.
• Group 2: Voters who know something about the candidates and the issues but don’t think they have enough information to make a worthwhile decision, and
• Group 3: Voters who think they are informed enough to make an educated decision.
The people who know almost nothing are doing neither themselves nor the community any favors by going to the polls today.
But most of the people reading this editorial are either in Group 2 or Group 3 — even though they may think they are in Group 1. They may not think they have enough information to make an informed decision, but those same studies suggest that they really do.
Whenever people use any decision-making factor — even something as simple as “when in doubt, vote anti-incumbent” — they have enough of an understanding of what’s at stake to participate competently. And even when people vote completely at random, the results usually cancel each other out.
I’m pleased to report that the editorial did a pretty good job of accurately summarizing some key findings from political science research over the past few decades. They’re right: partisanship is the key determinant of vote choice in U.S. elections. Democrats usually vote for Democrats and Republicans usually vote for Republicans.
They’re also right that people can and regularly do use shortcuts (“heuristics”) to help them make voting choices. Research in this area has shown that when people use shortcuts, they make a “correct” voting decision about 75% percent of the time. (In that study, voting “correctly” meant picking the same person using limited information as you would when you had access to full information about the candidates and campaign.) This finding is from Lau and Redlawsk 1997: http://www.uvm.edu/~dguber/POLS234/articles/lau.pdf
And finally, Page and Shapiro (1992) are to be credited with the “voting at random doesn’t matter because it tends to cancel out in the aggregate” argument:
I think the more interesting question, however, is the Press-Citizen’s argument that those who know nothing and completely guess (Group 1) should stay home and not vote. This is certainly somewhat of an elitist and anti-democratic argument to make, that only those who are well-informed should be voting. What do you think? Is it better for people to vote even if they have no idea what they’re voting about? Is a lower turn-out in an election a good thing if those who are voting are the ones who truly care?