“It seems that more than 96
percent of voters have already made up their minds about this
election,” the ad begins.

“Well, I guess some of us are just a little harder to
please,” it continues. “We’re not impressed by political spin
and 30-second sound bites. Before you get our vote, you’re going
to have to answer some questions. Questions like, ‘When is the
election?’ ‘How soon do we have to decide?’ ‘What are the names
of the two people running?”’

As you might have guessed, the ad is a spoof. We’re in that
blissful few months before an election in which NBC’s “Saturday
Night Live” becomes really, really good.

Even though the ad is an exaggeration, it’s not an outright
lie. This election will probably be decided by a tiny fraction
of the electorate in eight or nine states. The undecided voters
in those states are popularly portrayed as people who just can’t
make up their minds. But that’s not quite right. They aren’t so
much “undecided” as uninterested and, frankly, uninformed; in
political-science parlance — and SNL ads — they are “low
information” voters.

World Series

It’s worth stopping here to clarify something:
“uninformed” does not mean “dumb.” We’re all uninformed
about certain topics. You wouldn’t believe how little I know
about, say, baseball. I’m vaguely aware that it happens, and
that it culminates in a World Series, but I can’t tell you who
won last year, or who’s in contention this year. Baseball just
isn’t something I pay attention to.

Lynn Vavreck, a political scientist at the University of
California, Los Angeles, says that uninformed voters have
roughly the same relationship to politics that I have to

“They are lower on political information, for sure. That’s
a function of being not that interested and not paying
attention,” she said. “It’s not that they can’t comprehend the
information, or that they’re at a balancing point and can’t
decide. They’re just not dialed in. They’re not getting all the
information you or I are getting.”

Vavreck asked thousands of voters — both decided and
undecided — a battery of basic, multiple-choice questions about
who’s who in politics. The questions were designed to be easy.
You didn’t have to know that John Boehner is Speaker of the
House. You just had to know he is a congressman rather than a
judge or the vice president.

According to Vavreck’s polling, only 35 percent of
undecided voters could identify Boehner’s job as
“congressman.” Only 69 percent could say that Joe Biden is the
vice president rather than, say, a representative. Only 17
percent can identify Chief Justice John Roberts as a judge.

Decided voters have an easier time rattling off the job
titles of Boehner and Biden, as well as those of Harry Reid,
Eric Cantor, Mitch McConnell and Nancy Pelosi. (Interestingly,
they struggle more than undecideds to identify Roberts.)

That’s likely because decided voters are paying more
attention to the election. About 43 percent of decided voters
say they’re following the presidential election “very
closely.” Only 12 percent of undecided voters say the same.

Recognizing that undecided voters are mostly uninterested
voters helps to clarify the trajectory of the presidential
campaign. In their book “The Timeline of Presidential
,” Robert Erikson and Christopher Wlezien show that
voter preferences tend to be very stable in the fall, but that
campaign observers — the authors analyze people betting money
in online political prediction markets — tend to assume those
preferences are far more volatile.

Psychological Projection

The misjudgment makes sense as an act of psychological
projection. To people personally invested in politics, the
homestretch of the campaign appears loaded with the kind of
political information that could change voter opinions. There
are debates, a flood of ads, inevitable gaffes, the crush of
election news — maybe even an October surprise or two.

But undecided voters are precisely those least likely to
tune in to the debates, which helps explain why debates
typically have little effect on elections. They’re the least
likely to care about a gaffe — or even to know when one has
occurred. They’re more likely to throw out political mail and
tune out political ads. If they live in a swing state, they’ve
already been buffeted by — and proved immune to — months of
commercials and phone messages.

Vavreck has been tracking a group of 44,000 voters since
December 2011. When she started, 94 percent were already leaning
toward a candidate. Of the 6 percent who were truly undecided,
33 percent now say they’re going with Mitt Romney and 37 percent
with President Barack Obama. The ranks of the original undecided
voters were partially replenished by voters who had expressed a
preference in 2011 but have since grown uncertain. Of the new
undecideds, slightly more were Romney supporters in 2011 than
were Obama supporters, but the total numbers are small.

There’s little reason to believe that undecided voters in
this campaign will break sharply toward one candidate. The votes
of the undecideds seem to be roughly evenly split, and if any
big news happens between now and the election, they’re likely to
be the last to know about it, and the least interested in
following up on it. If Obama is going to turn this into a rout,
or if Romney is to salvage a win, it will probably require
changing minds that are already made up, or increasing (or
suppressing) turnout among base voters.

In other words, don’t expect the votes of the mythical
undecideds to actually be decisive. It’s likely to be the
decided who will, well, decide.

Our ideas can save democracy... But we need your help! Donate Now!

Ezra Klein

Follow Ezra on Twitter @ezraklein. Ezra Klein is the founder and editor-in-chief of Vox.com.