For years, the best tool to predict which party will gain House seats in any given election has been the so-called "generic ballot test." The "generic" is a poll question that asks voters which party they’d support in the upcoming congressional election. While it can’t tell us exactly how many seats one party or the other should expect to gain/lose, it does give us a good idea of the range we can expect. And, right now, Democrats should be very happy about what they are seeing. The RealClear Politics average shows Democrats with a whopping 10.5 percent lead on the generic. That comes close to the 11.5 percent lead Democrats held in the late fall of 2006. That November they picked up 31 seats and control of Congress.
However, with just over a year—and roughly 300,000 news cycles—before the midterms, is it too early to pay attention to what the generic ballot is telling us? And, given the structural advantages held by Republicans, just how big of a lead do Democrats need to have for us to conclude that they have a good chance of flipping the House?
The conventional wisdom, which was supported by most pollsters I spoke with this week, says that the generic ballot isn’t meaningful until at least the summer of the election year. Others say we should only really start to pay attention to the generic as a predictor of seats gained/lost around Labor Day of the election year.
Recently, a long-time Democratic strategist sent me a chart showing the average generic ballot advantage in the year before the last three midterm elections. In the email message he sent along with the chart he wrote "Too Early." For example, about 400 days out from the 2010 election, Democrats had a five point lead on the generic. By the week before the election, Republicans had taken a substantial 9 point lead. In the end, Republicans won the House vote by six points—and gained 63 seats. In 2014, Democrats had a one-point lead (D+1) in mid 2013. By the fall of 2014, Republicans had a 2 point lead (a swing of four points) and ultimately won the national House vote by 5 points.
But, FiveThirtyEight’s Harry Enten argues that even a year out, the generic ballot question is actually quite predictive.
The generic ballot, even this early in a midterm cycle, can be quite predictive of the outcome of the following year’s House elections. Once you control for which party is in the White House, the generic ballot about 18 months before a midterm election is strongly correlated (+.78) with the eventual House result—i.e., the share of votes cast for the president’s party versus the share of votes cast for the opposition party.
In other words, there is a 78 percent chance that Democrats 7 point lead on the generic ballot this summer will translate into a seven point margin of victory in the national House vote next fall.
However, both may be correct. One Democratic pollster, who has been digging into this question of the predictive value of the generic ballot found that the generic is "directly tied to the incumbent president’s approval rating, which tends not to settle in for the mid-term" until about Labor Day of the election year. So, this pollster adds, "if we think Trump is likely to be more popular a year from now, then the current margin in the generic is misleading. If we think it’s likely to be the same or even lower, then the generic today could well be the same a year from now. Republicans may consolidate a bit around their congressional candidates, but that would still put them in the low 40s which will be really bad by historical margins. If Trump’s approval stays mired where it is, I think a generic in the range it is today is very possible.”
For example, in the fall of 2009, Obama’s job approval averaged 53 percent (and the Democratic generic advantage that fall was D+5). But, by the fall of 2010, Obama’s approval rating had fallen to 45 percent. And, the generic ballot moved dramatically toward the GOP. Contrast this with the 2006 midterm where George W. Bush’s overall approval ratings were weak in 2005 and stayed equally anemic in 2006. In the fall of 2005, Bush’s approval rating averaged around 41 percent. By the fall of 2006 it was hovering at 37 percent. As such, the D+6 generic ballot in the fall of 2005 turned out to be predictive of where the Democrats would ultimately end up a year later.
There’s also the question of whether a national generic ballot number means all that much at a time when there are few swing seats, more straight party voting, deeper partisan affiliation and more sophisticated gerrymandered seats. In other words, a generic advantage is just that, "generic." It misses the nuances and uniqueness of each individual district. The pollsters at the NBC/Wall Street Journal poll have found a way to address that. They looked at the generic ballot responses by congressional district.
In the most recent October survey, they found that Republicans had a six-point advantage in GOP-held seats (R+6), while Democrats had a 29 point advantage in seats they hold (D+29). What’s significant—or what NBC/Wall Street Journal pollster Bill McInturff called a “flashing yellow light,” was that the GOP advantage in seats they already hold dropped eight points from September to October—from R+14 to R+5. It also stands in stark contrast to the average generic advantage Republicans had in seats they held in the most recent mid-term elections (R+15 in 2010 and R+18 in 2014). Why does this matter? It is a sign that the drag we are seeing on Republicans nationally is taking a toll even in districts they already hold.
Finally, what is the “magic number” Democrats need to hit in the generic ballot? My colleague David Wasserman thinks Democrats need to win the national House vote by seven to eight points in order to flip the 24 seats they need to take control of Congress. Given that FiveThirtyEight’s Enten notes that the “true margin of error for generic ballot polls is about +/- 5 percentage points, even for those done at the end of the campaign,” Democrats will want to see a generic ballot advantage of nine to twelve points to feel confident that they can hit seven points on election day.
President Trump is most certainly his own brand. He has defined himself as someone who is both of the GOP and yet outside of it. Even so, there are growing signs that the fate of House Republican control is directly tied to Trump. As his approval rating sinks, and divisions within the party grow, the GOP deficit in the generic congressional ballot grows as well. Individual GOP candidates are already doing all they can to insulate and localize their races. But, in a wave year, it is the national mood - not the individual campaigns—that is the most determinative. The good news for Republicans—there still is time for Trump to turn his sagging numbers around—and as such help narrow the Democrats generic ballot advantage. The bad news: his desire to play to the base limits any opportunity to move his approval ratings much above 40 percent.