Even without a winner declared Tuesday night, it was clear based on the results what the race would tell us about the 2018 midterm elections.
Democrat Conor Lamb holds a narrow lead over Republican Rick Saccone, but it really shouldn’t be that close: Donald Trump won by 20 percentage points by 2016 and the district has been solidly Republican for a little over a decade.
But it’s not just about this one race in isolation. It’s about what all the signs are telling us.
Yes, dynamics may change in the lead up to November. Events can change the course of history.
But at the present time, Lamb’s performance in Pennsylvania 18 is merely the latest sign Democrats are surging right now, spelling trouble for Republicans heading into the midterm elections.
The special elections
As I pointed out on Monday, the federal special election held before Pennsylvania 18 showed Republicans in a poor position. In the average of seven special elections before this one, Democrats were outperforming their partisan baseline (based off the previous two presidential results in the district) by 16 percentage points. In Pennsylvania 18, Lamb, the Democrat, outperformed it by 22 percentage points – a little better than average and essentially matching what they did in the Kansas 4 special election in April 2017.
The overperformance in special elections by Democrats is key to understanding the national environment heading into the midterms. When parties do well in special elections, they usually do well in the midterms. When they do poorly in special elections, they usually do poorly in the midterms.
Going back to 1994, there has been no cycle in which Democrats were doing this well in special elections. The only cycle that was close was 2006, when Democrats would go on to win 30 seats and control of the House. In both 2010 and 2014, Republicans were the ones who were outperforming their baseline in special elections.
If the past relationship between special election results and the midterm results holds in 2018, Democrats are in a strong position to take back the House in 2018.
Trump’s approval rating
Trump’s level of popularity or lack thereof also suggests Republicans are on their way to losing the House. Depending on the poll or average you examine, Trump’s approval rating is within a few points of 40%. The latest Gallup poll for example has him at just a 39% approval rating. That’s below where President Barack Obama’s approval rating was heading into the 2010 and 2014 midterms.
I collected data on each president’s approval rating at this point before each midterm election in Gallup surveys since 1946. Trump’s 39% approval rating at this point projects that the Republicans will lose the House popular vote by 10 points. (There’s a fairly wide margin of error in this estimate given that we are still have a lot time before the midterm.)
If the 10-point estimate is right on, Republicans will probably have a net loss of between 30 and 40 seats. That is far more than the net 23 seats they need to lose (if Lamb wins) for them to lose the House.
The generic ballot
This simple question usually asks voters whether they will vote for the Democratic Party or Republican Party in the midterm elections. The Republican Party remains in the worst position on the generic ballot for any majority party since at least 1938.
Their average 10-point deficit in live interview polls since February likewise points to them losing between 30 and 40 seats. And while there’s a definite potential for things to change, shifts on the generic ballot usually go against the president’s party (i.e. the Republicans in this situation).
Another worrisome sign for Republicans is there’s no indication that they will gain ground when pollsters switch from registered to “likely” voters. Beyond the enthusiasm that Democratic voters have shown in special elections so far, numerous generic ballot estimates have suggested that the results among registered voters (which most pollsters are surveying right now) may actually be underestimating Democrats.
In the George Washington University Battleground poll released this week, for example, the Democratic edge rose from 9 points among all respondents to 12 points among those who are “extremely likely to vote.” In a GWU poll taken in March 2014, Republicans were the ones who gained when the pollster went from all respondents to “extremely likely to vote.” This switch is in line with turnout statistics from prior midterms since 1978. Those midterms suggest that when there is a Republican president, Republican turnout is far lower than when there is a Democratic president.
The Pennsylvania 18 result is just another example of the enthusiasm of Democratic voters and the swing toward the Democrats among those in the middle of the electorate.