Two months ago, we were deluged with seemingly dismal news for Democrats.
Newsweek’s headline asked, “Are Democrats Losing Their 2018 Midterm Advantage?” If the question itself left anyone in doubt, the headline went on to offer an answer: “A New Poll Says Yes.”
The current news isn’t getting quite the same headline treatment, but it’s much more positive for Democrats, much more consistent with history and much less cherry-picked.
In May, politicos were transfixed by a few polls from Ipsos and CNN showing Democrats’ lead in the generic vote for Congress had dwindled to 1 or 2 points. One survey even showed the GOP ahead by 2 points.
There was plenty of reason for skepticism even then.
For example, other polls produced rather different results.
Surveys from YouGov, Morning Consult, Quinnipiac, IDB/TIPP, Fox News, the NBC/Wall Street Journal and others pegged Democrats’ advantage at 5 to 10 points.
To be sure, the polls on the lower end pulled down the averages — as low as 4 points in Democrats’ favor on one day.
Another cause for caution in interpreting the negative news were the surprising fluctuations in the polls that started the stampede toward downgrading Democrats’ prospects.
Ipsos’s May 16-20 survey gave Republicans a 2-point lead. Just days later, their May 21-25 survey put Democrats 5 points ahead, while their next survey, May 26-30, showed Democrats with a 6-point lead.
Professor Charles Franklin, who took an even more fine-grained look at the Ipsos data, found them showing a 4.2-point Democratic lead on May 13. Just five days later, on May 18, the GOP held a 5.6-point advantage, for a net shift of almost 10 points in five days.
I will tell you one thing that did not happen in the real world: the generic vote did not shift by 10, or even by 7 or 8 points in a matter of five days.
Franklin identified the likely culprit as well. Ipsos’s sample swung a net of 7 points toward the GOP in party identification, one factor which drives the generic vote. A more Republican sample is going to find more Republican generic voters.
But the story doesn’t end here.
As I noted earlier here, history shows that, as we get closer to Election Day, the generic vote gravitates away from the party holding the White House and toward the opposition.
In May of 2010, the generic vote was essentially a tie. By November, it moved to nearly 7 points in the GOP’s favor.
Similarly, in 2014, the generic was tied in May, but Republicans led by 6 points in the end.
That same pattern has maintained this year. While Democrats led by just 4 toward the end of May in the FiveThirtyEight average, the numbers have drifted toward the Democrats, who now lead by 7.5.
Based on historical trends, that generic vote margin today should translate into just under a 9-point national House margin on Election Day.
Of course, there is a substantial margin of error on that estimate.
But even more chilling for Democrats is this: For reasons I described here back in February (“Of Waves and Walls,” Feb. 28, 2018), it is possible for Democrats to win the national House vote by 7-10 points and still not take control of the chamber.
We’ve got a long ride until Election Day and no outcome is guaranteed. But cherry-picking polls, ignoring the fundamentals and focusing on apparent micro-trends won’t get us any closer to an accurate predication of what will happen.
Mellman is President of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House Members. Mellman served as pollster to Senate Democratic Leaders for over 20 years and as president of the American Association of Political Consultants.