The good thing about the Electoral College—our strangely still-surviving 18th-century experiment in federalism—is that it’s clear, coherent, and -commonsensical. If you live in Ohio, say, a state that’s closely contested in the presidential race this year, you know down in your bones that your ballot is important. More important, certainly, than the ballots coming out of Illinois, where President Obama remains so messianic that even the dead will rise up to vote for him, if necessary.
The bad thing about the Electoral College, unfortunately, springs from the same root: It’s clear, coherent, and commonsensical. The rise of personal computers over the last 30 years has taught everyone in America the deep truth of information theory: If it’s a system, it’s got inequalities. If it uses rules, it can be gamed. Added up one way, a voter in Wyoming casts what is easily the most influential ballot in this year’s presidential race: 3.52 times the value of a Texas ballot. Added up another way, people in the District of Columbia have the most say in the national election: 3.60 times more than voters in Wisconsin.
The really interesting numbers, however, are found when you try to estimate the worth of votes in the narrow margins of tossup states. Given recent polls, a single presidential ballot in Utah counts for very little—while each ballot in neighboring Colorado could influence the national result. If you live in Massachusetts, you probably won’t be missed if you stay home on Election Day. If you live in Nevada, you should crawl across the desert over broken glass to get to a ballot box.
What produces these variations are differences in voting-age population, local patterns of voter turnout, and the size of the gap between the candidates. Behind them all, lies the Electoral College: In any given election, voters in some states will contribute far more to the choice of president than voters in other states.
The Electoral College’s numbers—its 538 members—are determined by adding up each state’s senators and representatives. Every state has two senators and at least one member of Congress. One result is that no state can fall below three votes (the District of Columbia is also granted three votes), which means that small-population states are more valuable in relative terms. Another result is that larger states, commanding larger vote totals, are more valuable in absolute terms.
Thus, for example, the most consequential possible vote in a presidential election—we’re talking -theoretically, here; a logical fantasy—would be if only a single Californian voted: That one voter would direct 55 votes in the Electoral College—20 percent of what a candidate needs to win the White House.
But even when we bring the possible results a little closer to reality, there’s still a good-sized gap between the states. As a function of general population, Wyoming offers by far the most valuable ballots: Under 600,000 people with 3 electoral votes works out to 5.32 electoral votes per million. The District of Columbia, Vermont, North Dakota, and Alaska come next, each of them over 4 electoral votes per million residents. California stumbles down to last, with 1.48 electoral votes per million, followed closely by New York and Texas.
By that population measure, a Wyoming ballot is worth around three and a half times a California ballot. Of course, the effect is often minor: The 10 least populous jurisdictions average 4.1 electoral votes per million residents, but they add up to only 32 electors. The 10 most populous states average 1.6 electors per million but produce 256 electoral votes: a good way toward the 270 needed for victory. The 12 largest states, by themselves, could elect a president. Working from the other end, a candidate would need all 40 of the smallest states plus the District of Columbia to reach 270 electoral votes.
If we look not at total population but at the percentage that actually turns out to vote (averaged over the last three presidential cycles), the District of Columbia leaps far into
the lead, with 12.06 electors per million who vote. Wyoming at 11.32, Vermont at 9.30, and North Dakota at 9.19 round out the valuable end. Coming in last are Ohio with 3.31 electors per million of turnout, Wisconsin with 3.35, and Michigan with 3.40.
None of this is new, exactly. Back in 1968, J. F. Banzhaf published “One Man, 3.312 Votes,” which remains one of the most cited essays about the effect of the Electoral College. Curiously, if an argument against the American way of elections lurks in Banzhaf’s “power index” of voting, it derives from his conclusion that disproportionate electoral power actually belongs to voters in large states rather than small states.
The math of such statistical analyses is complicated and much disputed, but at the root lies the fairly simple fact that when a state with a large population is closely contested, it can multiply its voters’ electoral power: up to 3.312 times in a random election, Banzhaf claimed.
This electoral power lives in the voting gap—the number of decisive voters—between the candidates. If one candidate gains 3 million votes in a state, while the other candidate receives only 2 million, then the race could be said to boil down to the one million who made a difference.
So, for example, the District of Columbia ought to be at the top
of Electoral College power per voter. But given the tightness of the race between Mitt Romney and Barack Obama nationally, the District actually comes in dead last in power: likely to produce this year’s least consequential votes. With Obama favored by over 90 percent of its voters, Washington commands only 14.04 Electoral College votes per million
decisive voters. Utah at 14.46 electors per million deciding votes, New York at 14.57, and Oklahoma at 16.32 are similarly low.
Compare that with Iowa, Florida, Wisconsin, and Ohio, all at over a hundred electors per million decisive voters. And then the big four: Nevada at 399.12, Colorado at 562.47, New Hampshire at 836.06, and, overwhelmingly, Virginia at 932.85. In terms of producing an Electoral College vote, given the current state of the polls, a vote in Arlington, Virginia, is 66 times more valuable than a vote across the river in Washington, D.C.
Right in the middle, at 38.79 Electoral College votes per million poll-gap voters, lies Maine. Which is where it should be, one supposes: As I said, the Electoral College matches surprisingly well our commonsense understanding of the nation.
Our commonsense understanding of voting, for that matter. Little states concentrate their ballots enormously but end up with low influence on the election. Large states dilute their ballots but generally determine the outcome. States dominated by a single candidate tend to diminish the worth of their individual votes, while battleground states tend to magnify their votes. And the Electoral College—our strangely still-surviving 18th-century experiment in federalism—works to smooth the whole thing out.
Joseph Bottum is a contributing editor to The Weekly Standard and author of The Christmas Plains.