Vermont's Bold, Republican-Led Experiment in Deciding Elections by Land Area Instead of Votes

A losing candidate tries to win 2014's last election.
Photographer: Alex Wong/Getty Images

Shortly after 10 a.m. Eastern time Thursday, Vermont's legislature will meet to elect the state's governor. Two-term Democratic Governor Peter Shumlin secured a narrow win in November 2014's popular vote; unfortunately, if no Vermont gubernatorial candidate takes more than 50 percent of the vote on election day, 180 lawmakers get to pick the winner. This has happened 24 previous times in the history of Vermont, which really should tell them something.

But before they can change that, they have to deal with Shumlin. He won 46.4 percent of the vote in November, while Republican Scott Milne won 45.1 percent. Six other governors are taking office this month having failed to win more than 50 percent of the vote, including Shumlin's possible neighbor, Massachusetts's Charlie Baker. Only Shumlin might end up losing to a guy who won fewer votes. That's because Milne has kept on campaigning, asking legislators not to vote for the popular vote winner, but to vote for the winner in their districts. Republicans hold only 48 of the House's 150 seats, and seven of the Senate's 30 seats, but thanks in part to spoiler candidates, Milne won a majority or plurality in 91 of the 180 districts. (Devil's advocate: The presence of a Libertarian candidate probably hurt Milne.)

Milne has turned this trivia into a cause. "In the same way the U.S. electoral college guarantees Vermont a say in national elections," wrote a Milne supporter in the Burlington Free Press this week, "our state Constitution gives Vermonters a voice where none would otherwise exist, even when a so-called 'one-vote' system points to a win for Peter Shumlin." (The "so-called" was a nice touch.) A mysterious group called Vermonters for Honest Government is even running TV ads, asking voters to contact legislators and back Milne.

Milne's advantage should be familiar to the Democrats who've been fending off swing state Republican plans to divvy up electoral votes by congressional district. The modern Democrats play better with human voters than they play with land area. Thanks to the urban strength of the party and to gerrymandering maps that have reinforced it, Democrats have lost most of the country's 435 House districts in all but two elections since 1994. Since 2000, the Democrats have won the popular vote for the presidency three times. But if the elections were decided by the House and Senate, and each legislator voted the way his constituents voted for president, only once—in 2008—would the Democrat have taken the White House.

The 2000 election's obvious enough—our Earth's version of Al Gore won the popular vote but lost anyway—so start the clock in 2004. That presidential race was uncomfortably close for George W. Bush; the incumbent president won 286 electoral votes, the closest re-election campaign since Woodrow Wilson's in 1916. And had the race been thrown to the congressional districts and states, it would have been a little more of a walk. The president won 255 districts and 31 states. In the Milne system, with every senator voting as his state votes, that's good enoug, for 317 electoral votes. Bush won the 2004 popular vote by 2.4 percent; he won the fantasy electoral college by 18 percent.

In 2008, the popular vote winner would have triumphed again. In our reality, Barack Obama won 28 states (plus D.C.) and 242 congressional districts. Arizona Senator John McCain won 193 districts and 22 states. The real-world electoral college gave Obama a 365-173 victory; under the Milne system, Obama would have won with 301 electoral votes to McCain's 237. Obama won the 2008 popular vote by 7.2 percent, but he won the Milne College by 12 percent.

On to 2012—the first election after a round of gerrymanders that generally favored Republicans. That year, Mitt Romney won 226 congressional districts to Obama's 209. Obama won 26 states and D.C., but his weakness in rural districts and the windy new Midwestern districts drawn to vote in Republicans would have led to a Milne College debacle. Romney, having lost the popular vote by 3.9 percent, would have won 274-264 in legislative votes.

This is all wildly counterfactual, and any time somebody posits an election that isn't decided by our Electoral College, somebody else points out that candidates would campaign differently if they weren't incentivized to win states. They'd run different media buys, stump in different areas, turn out different voters.

That's probably what Peter Shumlin would have done to, if 91 Vermont legislators decide that they should overturn November's popular vote and go with whoever won their district.