Early last month, The Washington Post released polls from all 50 states, conducted in partnership with SurveyMonkey. There was a startling result in the data: Democratic presidential nominee Hillary Clinton led GOP nominee Donald Trump by 1 percent in Texas. That's within the margin of error, but given that Texas has voted for the Republican by an average of 15 percentage points since 1980, a Democrat even being within shouting distance of the Republican in the state was sort of hard to believe. I will admit that I was skeptical.
That poll was conducted over the month of August, when Clinton's lead over Trump nationally was at its peak. It would be hard for Clinton to have a massive lead across the country and be trailing Trump by a wide margin in Texas, given that Texas comprises 8 percent of the U.S. population. So if Clinton was doing much better than Trump everywhere, it would make sense that she was doing better than past Democrats in Texas, too.
On Tuesday, The Post released new data from a slew of battleground states. Trump leads by 2 points in Texas. There was also a poll released on Tuesday from the University of Houston. It showed Trump up by 3.
These new polls come out when Clinton is near another peak in her national lead, so the same arguments apply. But the question remains: Could Clinton actually win the state?
Well, sure. She could. There hasn't been a whole lot of polling in Texas, presumably in part because it's been assumed that Clinton wasn't going to compete there. But the current polling average calculated by RealClearPolitics has Trump with a 5.7-point lead — about where his lead is in Georgia, which is consistently discussed as a possible battleground state. (The RCP average doesn't include the new SurveyMonkey poll or Houston's.) His lead isn't much wider than the gap in Iowa has been in recent weeks, either.
Notice that most of the states on the graph above rise and fall with the national polling average. That's in part because there have been more polls in those states. There have been eight polls included in the RCP average from Georgia since the beginning of September, and only four in the average for Texas. (In our September 50-state poll, Georgia was tied. Now, Clinton leads by 4 points.)
There's no consistent link between national polls and state polls, though. Over the past two months, Georgia's been an average of 7 points friendlier to Trump than the national average. In recent polls, that gap has widened as Clinton's done better nationally and polls have shown Trump doing consistently well in the Peach State. In Ohio, voters have been an average of 3 points more friendly to Trump than the national average. When the national polling picture has been stronger for Clinton (as it was over the past few weeks), that was enough to pull Ohio into her column. The link between national polls and state polls exists, but it's not entirely predictive.
That it's close at all is probably a function of the candidates. Texas backed home-state senator Ted Cruz in the primaries by a wide margin; Trump's brand of Republicanism isn't entirely in line with the state's. Compare with four years ago. In mid-October 2012, Mitt Romney had a 17-point lead over President Obama in Texas, according to the RCP average — three times the margin as it stands now.
There are reasons to think that Clinton's got a better shot in Texas than, say, Wyoming, including that it is majority-minority and 40 percent of its residents are Hispanic. (Clinton leads by 50 points with Hispanics.) Wyoming is 84 percent white; Georgia is 54 percent white but 32 percent black. Her campaign plans to press the issue, running ads in the state for one of the final three weeks.
All of this is a hopelessly complicated way of saying something obvious: If Clinton beats Trump by a wide enough margin overall, that means she's pulled a number of states to the left. And it means that, yes, she could win Texas. If Clinton wins narrowly? Probably not.
Not super insightful. But: Let's start polling in Texas as much as we do in Georgia, regardless.