Washington Post poll indicates unexpected states may be in play

♦By Rich Shumate, Chickenfriedpolitics.com editor

(CFP) — When it comes to wondering which way Southern states might fall in this year’s presidential election, conventional wisdom holds that, outside of a few states along the Atlantic seaboard, Donald Trump doesn’t have a thing to worry about.

A paucity of public polling in most of the South also means that we don’t have any way to judge whether this conventional wisdom is wise. But now, the Washington Post and Survey Monkey have come along with an unusual poll of all 50 states that offers some incredible results for a number of places in the South.

In fact, should these results be borne out in November, incredible won’t be a sufficiently strong adjective.

In eight of the 14 Southern states, the poll found that Trump’s lead over Hillary Clinton was only in the single digits—not just in battleground states such as Florida, Georgia, North Carolina and Virginia but in some rock-ribbed GOP states such as Arkansas, Mississippi, South Carolina and Texas.

In Arkansas, which Mitt Romney carried by 23 points in 2012, Trump led Clinton by just 9 points in the poll; in South Carolina, by just 7. And in Mississippi—a state that hasn’t gone Democratic since 1980—Trump’s lead over Clinton was a mere 3 points. No, that was not a misprint—3 points in Mississippi.

In perhaps the biggest shock of all, the poll found that Texas, with its bounty of 38 electoral votes, was a flat-out tie, with both candidates at 40 percent in a state a Democrat hasn’t carried since 1976. (How long ago was that? Donald Trump was still a year away from marrying wife No. 1, Ivana.)

Of course, one must greet the results of any poll showing Trump in possible trouble in Mississippi, Texas and Arkansas with a considerable dollop of skepticism. Casting more doubt on these findings is the fact that this poll was an online survey conducted on Survey Monkey’s platform that did not employ a random sample and, thus, a margin of error can’t be calculated.

However, despite those shortcomings, what makes this poll noteworthy is its extremely large sample size — 74,000 registered voters, about 70 times the size of a typical poll — which could compensate for some of the measurement and sampling issues posed by a non-random sample.

Also, battleground state results in this poll are consistent with other public polling in those states. (In Florida, for example, the poll showed Clinton with a 2-point lead, which is close to what the other polls have found.) And there is no reason to believe this methodology is inherently more inaccurate in non-battleground states.

So what might explain these wacky results that so assault our conventional wisdom? The answer could lie in Trump’s pronounced unpopularity with minority voters.

In Mississippi, for instance, the black voting age population is about 35 percent, and blacks register and vote at rates comparable to whites. So if Clinton is taking almost all of the black vote, she would need to capture just 25 percent of the white vote to win. True, President Obama couldn’t pull that off against Romney four years ago, but, then again, the GOP was much more united behind Romney than it is behind Trump.

And in Texas, blacks and Latinos make up an even larger percentage of the voting age population, 44 percent. If Clinton can galvanize those minority voters in large numbers, as little as 20 percent of the Lone Star State’s white, non-Latino vote could be enough to win (that’s a conservative estimate taking into account the fact that Latino voters lag black and white voters in their registration and participation rates.)

Florida, where Clinton led in the poll, has a minority voting age population of 31 percent, and, unlike in Texas, Latino voters in the Sunshine State register and turnout to vote at rates comparable to white voters. So with just 30 percent of the white, non-Latino vote, she could win.

A look at polls results those Southern states where Trump is doing well also supports the premise that minority antipathy may be what’s bedeviling him in the region. He was winning by more than 20 points in Kentucky, Tennessee, Oklahoma and West Virginia, all states with relatively small minority populations. But in three of the five states where the black voting age population is more than 25 percent, his lead was in single digits.

The only states where the correlation between high minority population and a reduced level of Trump’s level of support didn’t hold up were Alabama (where Trump led by 26 points) and Louisiana (where he led by 16.)

Arkansas—where the black voting age population is only 15 percent and he was only ahead by 9 points—stands out as an anomaly against the premise that a paucity of minority voters leads to wide margins for Trump. Then again, Hillary Clinton did live in Arkansas for nearly 20 years before decamping for New York, and her husband was elected governor of the Natural State five times. Perhaps they have more residual political roots than Republicans are anticipating.

So should we all put some money down on a Clinton victory in Arkansas, Mississippi or Texas? Despite the intriguing results from this poll, the odds of that happening would still have to be considered rather long. But this has been a political year where the only thing we’ve been able to expect is the unexpected. So it is not inconceivable that, if Clinton can galvanize the region’s minority vote and Trump runs behind Romney, election night might turn out more interesting than usual across the South.