Analysis: The Presidential Race Is Close. Get Used To It
The expected Clinton-Trump presidential race is no blowout. It may well be one by the time November rolls around, but right now the polls show a close race that is following similar patterns to recent elections.
In this respect the polls are upending conventional wisdom as the general election phase approaches, just as they did throughout the primaries. Pundits, including data-driven ones like Nate Silver, didn’t think Donald Trump would be the GOP nominee, despite polls showing him ahead of the pack for months. With the primary mostly in the rear view mirror, already some are predicting that Hillary Clinton will easily cruise to victory over Trump in November.
But the polls, again, are saying not so fast.
Twitter exploded Tuesday morning as Quinnipiac University released polls from three crucial swing states showing Clinton and Trump locked in closer-than-expected races. There was some criticism about the makeup of who was sampled, and questions as to whether the Quinnipiac polls were outliers. But the Quinnipiac polls were not alone in showing the fall election as a real contest. Just on Wednesday, a national Reuters poll was released, showing Clinton with just a 1-point edge.
Looking at the body of polls released since Trump wrapped up the nomination suggests the race is both competitive and following familiar state by state contours. The margins from recent polls pitting Trump and Clinton head to head offer unmistakable echoes of 2012. A Dartmouth College poll of New Hampshire pegged the race there at a 5-point lead for Clinton, similar to the 6-point margin for Obama in 2012. The Boston Globe/Suffolk University poll of Massachusetts gave Clinton a 24-point lead, nearly identical to Obama’s 23-point margin over Mitt Romney. Polls this week in Florida, New York, West Virginia and Louisiana were also within a point or less of the 2012 margin.
Comparing the 10 most recently polled states to the final 2012 margin, four are more favorable to Democrats, four more favorable to Republicans, and two are identical to the 2012 margin.
There are always possible criticisms to make of individual polls. Dig deep enough into the crosstabs and the demographics, and you can find reason to cast doubt on any poll result. One could argue the Dartmouth poll may have polled too many men, Quinnipiac may have polled too many white voters; Public Policy Polling may be skewed from under-sampling cellphone users; and the Reuters poll shifted too quickly to be reliable. So yes, it can all be explained away, if that's your aim.
But a big reason to aggregate polls is to avoid reading too much into the minutiae of any one poll, and to counter-balance, skew or error from a specific poll. And in this case, they are all pointing to roughly the same current state of play. Whether Trump is down 1 or 4 points, or whether any of these smaller errors might shift things 2 or 3 points in a given state, are both pretty far beside the point. Unless all of the polls are systematically missing in the same direction, the current state of play is a close race that pretty much follows the lines of previous contests.
To be clear, if an electoral repeat is in the offing, the (narrow) advantage would still be given to Clinton. The 2012 map to which these polls are hewing closely was, after all, an Electoral College win for the Democrats. The electoral map is already unfavorable to Republicans, and inexorable demographic changes are peeling more and more of the electorate away from them. Trump remains incredibly unpopular with voting blocs he would need to remake the map, and so far he has shown little interest in changing his tune to appeal to them.
Unless he does, and unless some other major event shakes up the race, the fundamentals suggest the race should remain tilted in Clinton’s favor. But right now, both candidates have just left the starting blocks, and polls show it very close.
Steve Koczela is president of the MassINC Polling Group and a regular contributor to WBUR Politicker. He tweets @skoczela.