This Weird Election Is Scrambling Some Reliably Red and Blue States

All things seem possible this year. Photo: Justin Sullivan/Getty Images

Could Hillary Clinton carry Arizona, Georgia, and Texas (!) in November while losing Iowa and Ohio? That is a distinct possibility, according to a vast new survey of the 50 states published by the Washington Post in conjunction with the online pollster SurveyMonkey, based on surveys conducted over much of August. And while this kind of poll should probably not be taken to the bank in terms of specific accuracy, it does show the changing nature of the presidential battleground pretty dramatically, and in ways that echo other evidence emerging this year.

The big development the Post survey confirms is the power of race and educational levels to change the map. Trump is doing exceptionally well in states with relatively low minority populations and relatively high numbers of non-college-educated white voters, regardless of past partisan leanings. That helps explain why Iowa (the only battleground state in which Trump currently leads, according to the RealClearPolitics polling averages), Ohio, and Wisconsin are looking relatively good for the mogul, while Pennsylvania and Michigan are also within striking distance for the Republican. Obama carried all of these states twice; Michigan, Pennsylvania, and Wisconsin haven’t gone Republican since 1988.

States with a combination of a large African-American and/or Latino population and a relatively high percentage of college-educated white voters, however, are leaning the other way. The best example is Virginia, which went Republican in every presidential election from 1968 through 2004 before going twice for Obama and now leaning heavily (plus 8) to Clinton. North Carolina has followed a similar pattern, albeit not quite as strongly (it went Republican in every election from 1980 through 2004, and it was narrowly carried by Obama in 2008 and by Romney in 2012; the Post/SurveyMonkey now shows a tie there). And now Arizona (which has gone Democratic just once since 1948), Georgia (Republican since Clinton narrowly carried it in 1992), and Texas (last carried by a Democrat in 1976) are moving in the same direction for the same reasons.

While the Post headlines all of these findings as good news for Clinton given the overall standing of the candidates in the 50 states, they do illustrate Trump’s narrow path to victory: winning the previously Democratic states where he’s now leading or competitive (the latter includes Florida, a close state that does not fit any of the usual patterns, in part because of the unusual composition of its large Latino population) while holding off the Democratic trend in the red states where he is struggling.

As for the cognitive dissonance political junkies experience in looking at maps where Iowa is red and Texas is blue, well, we’ve been here before. New York and California were for many years the quintessential swing states, before the former became more or less reliably Democratic in the 1970s, while the latter turned red in the 1980s, and then blue in the 1990s. Illinois switched from red to blue at the same time, and to the same degree, as California. Kentucky was once the ultimate bellwether state; now it’s reliably Republican. Oregon was dead even in the close elections of 1976 and 2000; now it’s reliably Democratic.

The new map we are adjusting to this year is largely the product of demographic changes, but it could well be that the implications of these changes are being exacerbated by this particular presidential campaign. A future Republican candidate, for example, may sacrifice some of Trump’s appeal to white working-class voters in exchange for a fighting chance to win Latinos or college-educated white women. So don’t memorize the “new map” just yet.