It was President Obama's fault.

No, I haven't thrown in with the Obama-haters who pin blame for everything from Ebola to the polar vortex on the Democratic president.

I'm reporting the takeaway from a fascinating analysis of the 11 DFL defeats on Nov. 4 that flipped control of the Minnesota House to the GOP, giving that party a 72-62 edge in 2015-16. The analysis suggests that Obama's unpopularity had a lot to do with results in those 11 districts.

It's the handiwork of Brian Rice, the Minneapolis attorney/lobbyist/born-and-bred DFLer and election math whiz. He examined the House results in every presidential midterm election for the past 60 years — 16 in all — and scored what happened to the state House caucus allied with the president's political party in each of those elections. (Before 1974, those caucuses used the nonpartisan names Liberal and Conservative. Rice took those terms as proxies for Democrats and Republicans, which they were.)

In 14 of those 16 elections, the caucus associated with the president lost seats. The two exceptions were in 1990, when President Bush the First was basking in the glow of winning the Cold War and gearing up for the Gulf War; and in 2002, when President Bush the Second was on a post-9/11, pre-Iraq War footing and the DFL was in disarray after the death of U.S. Sen. Paul Wellstone. The GOP gained one House seat in 1990 and 12 in 2002.

Cast aside those two exceptions, and consider the number of state House seats typically lost by the president's party in midterm elections. They ranged from 6 to 37. The average: 17.

By those lights, this year's swing of 11 seats doesn't seem so stunning. Neither does the fact that 10 of the 11 districts that switched from DFL to GOP are in Greater Minnesota — not when one considers the president's popularity pattern in this state.

A Survey USA poll conducted Oct. 27-30 in Minnesota (which Obama carried in both 2008 and 2012) found that his approval statewide had slipped to 40 percent. That's well into the political peril zone.

More significant for the state House was the rural/urban split in voters' evaluation of the president. Obama is in just a little trouble with metro voters, with 46 percent approving and 48 percent disapproving his performance in the Survey USA poll. But he's toast in Greater Minnesota, with disapproval levels ranging from 58 percent in northeastern Minnesota to 70 percent in the western half of the state map.

Rice's conclusion: "A president's standing at midterm affects state House races in a major way. Obama was a big factor in this race. The dichotomy between Obama's approval rating in the metro area and Greater Minnesota is huge. Greater Minnesota wasn't shielded from the 'red tide' against Democrats and the president nationally. It participated in it."

So much for the late U.S. House Speaker Tip O'Neill's famous maxim "all politics is local." Not only was that not the case in Minnesota House elections this year. It probably never was.

Rice's data explain a lot about Minnesota's bifurcated election results and the campaign that led up to them. For example: Why did Republican U.S. Senate candidate Mike McFadden mention DFL Sen. Al Franken's association with Obama in nearly every public utterance for at least six months? McFadden undoubtedly knew how weak Obama had become outstate, and that for a well-heeled investment banker from Sunfish Lake to have a chance, he had to make Franken the functional equivalent of the president in outstate minds.

It didn't work, but not for lack of trying.

GOP awareness of antipathy for Obama could be why state House Republican candidates often made a point of linking the state's health insurance exchange, MNsure, to Obamacare, a scary word in a population that's disproportionately elderly. It may be why the year's hot new epithet in legislative races was "metro-centric," a vague term that in outstate ears evokes a connection to the diverse groups that comprised Obama's twice-winning coalition.

But the compelling data Rice assembled leave me wondering: If Republicans saw the Obama drag on DFL popularity outstate and exploited it, DFLers must have seen it, too. If they did, what did they offer in defense? DFLers might have pointed to the fact that a majority of the beneficiaries of the 2013 increase in local government aid and the 2014 increase in the minimum wage live in Greater Minnesota. They could have pounded the fact that Greater Minnesota projects comprised a 44 percent share of the 2014 bonding bill, compared with 31 percent for metro-specific projects. (I'm not counting the Capitol renovation among the metro projects. If any building belongs to the whole state, it does.)

Maybe a few words should have been uttered periodically in defense of a president who has overseen 56 consecutive months of job growth and extended the benefits of health insurance to millions who would not have otherwise have it — plenty of them in rural areas.

Maybe those were prominent DFL themes outstate. Maybe they could have been more prominent.

This much is clear: For at least a dozen years, DFLers have considered suburban districts their toughest and most telling battleground with Republicans. Few outstate districts were on the focus lists DFLers shared with journalists at midsummer.

This year's results suggest that the DFL's focus should have been wider. And Rice's finding that, whether the president is a Republican or a Democrat, presidential popularity at midterm is highly correlated with voters' choices as far down the ballot as the state House should factor into both parties' campaign strategies in 2018.

Lori Sturdevant is a Star Tribune editorial writer and columnist. She is at lsturdevant@startribune.com.