As election day quickly – nay, mercifully – approaches, commentators have been contemplating the possibility that Barack Obama might win the election while losing the popular vote, and fact that this would be due to the president’s marked shortage of supporters in “the South”. Ironically, if this should happen it might well give the Electoral College a new lease on life.

Earlier this month, Gallup released crosstabs showing that Obama was leading Mitt Romney in the East, Midwest, and West by four to six points, but losing to him by 22 points in the South. The president’s gap in Dixie – which Gallup defines as including Oklahoma, Texas, and Kentucky in addition to the usual suspects — was wide enough to give Mitt Romney an overall lead of 50-46 among likely voters.

If Gallup’s Oct 9-15 poll proves prophetic, Romney’s popular vote victory would be due to wide differences of opinion between the South and the rest of the country. This will perhaps put the Electoral College in a more flattering light for those who have been its greatest detractors in recent years: liberals upset over the results of the 2000 election, when George W. Bush won the White House without winning the popular vote.

But first, some words on the data.

Some commentators have argued that Gallup is likely overstating the case. Nate Cohn at TNR has noted that other national polls give Romney a narrower lead in the South – 11 points when averaged together, and including the Gallup poll. He points out that Obama has been doing more poorly in New England and the upper Midwest than he did four years ago against John McCain.

I’d argue that the polls are generally underestimating Romney’s support in the two most reliably conservative cultural regions of the country – the Deep South and Greater Appalachia – primarily because they fail to recognize the country’s actual cultural geography.

I traced the historical development and characteristics of our continent’s sections in my recent book, American Nations: A History of the Eleven Rival Regional Cultures of North America. Our real regional map — available for download here – doesn’t respect state or even international boundaries, and can’t be captured by state-level polling results. (For a brief synopsis of the nations, read this article from the magazine.)

The regional disparities in this election are actually muted by polls that lump southern Ohio, Illinois, Indiana, and Missouri – all part of Appalachia – in with “the North” or “the Midwest”. Similarly, Appalachian support for Romney is diluted by state-level polls that roll swaths of the region in with the Virginia and North Carolina Tidewater country. Deep Southern support is watered down by the inclusion of Spanish-founded, blue-trending regions in the southern parts of Texas and Florida. Were polls to be sorted by regional culture, the gap between the candidates would certainly be far wider.

Although a Yankee himself, Mitt Romney is crushing the president from Mississippi to West Virginia — states completely or overwhelmingly controlled by these two cultures – because voters lack a more conservative, libertarian-minded alternative. But there’s a catch: Appalachia extends into—but does not outright control — several key swing states including, most importantly, Ohio and Virginia. With the rest of Virginia (Tidewater) and Ohio (Yankeedom and the Midlands) trending toward Obama, Romney needs Appalachian voters to turn out at the polls in strong numbers to win those state’s electoral votes. Whether they do or not (given their lack of excitement about either candidate) is likely to decide the election.

If we do experience a split between the popular and Electoral College votes, it will indeed be largely due to Obama’s profound weakness in these two very red regions. It will also underscore the original underlying purpose of the College: to ensure a coalition of the states – and in 1789, those were nearly analogous to the regional cultures – would choose the chief executive, rather than a majority of the national population at large. It’s obviously an undemocratic way of doing things, but it’s an outgrowth of the fact that we’ve never been a nation-state like Japan or Denmark, but rather a federation of rival “nations” more akin to Canada or Belgium. If one is to legitimately represent the entire federation, the reasoning goes, they ought to have support across a broad range of regions.

If the results of next week’s vote prompt a discussion of how we elect the president, regionalism should be central to the discussion. Federalism does confound Democracy, but it does so with a purpose. Americans will have to decide if we have truly outlived that purpose or not.

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

Colin Woodard is the author of six books, including Union: The Struggle to Forge the Story of United States Nationhood. He is the director of the Nationhood Lab at Salve Regina University’s Pell Center for International Relations and Public Policy.